Objective To propose the Schwalbe line (SL) as a new anatomical landmark, independent of the scleral spur (SS) location, for assessing anterior chamber angle (ACA) width quantitatively with high-definition optical coherence tomography (HD-OCT).
Methods Study subjects underwent dark-room gonioscopy and HD-OCT in one randomly selected eye. The authors developed a computer-aided program to define two new quantitative parameters for assessing ACA width: Schwalbe line-angle opening distance (SL-AOD) measured at the SL, and Schwalbe line-trabecular–iris space area (SL-TISA) measured 500 μm from the SL. The associations between SL parameters, SS parameters and gonioscopic grading were evaluated.
Results Seventy-three (47 females, 26 males) subjects were recruited, the majority of whom were Chinese (89%). The authors excluded 29 images (19.9%) owing to poor image quality, leaving 117 HD-OCT images (65 nasal, 52 temporal) for analysis. SL and SS could be identified in 95% and 85% of quadrants respectively (p=0.035). SL-AOD and SL-TISA were significantly correlated with SS parameters (all r≥0.85) and gonioscopic grading (all r≥0.69). In eyes with closed angles (n=36), SL parameters showed strong correlations with gonioscopic grading (r ranged from 0.43 to 0.44).
Conclusions Novel angle parameters, based on SL as a landmark, may be useful to quantify ACA width and to assess for risk of angle closure.
- Angle-closure glaucoma
- anterior chamber angle
- optical coherence tomography
- Schwalbe line
- diagnostic tests/investigation
Statistics from Altmetric.com
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.
- Angle-closure glaucoma
- anterior chamber angle
- optical coherence tomography
- Schwalbe line
- diagnostic tests/investigation
Primary angle-closure glaucoma is a major cause of visual morbidity in Asia.1–5 As glaucomatous damage is irreversible, new methods for identifying angle closure accurately are desirable. Anterior segment optical coherence tomography (AS-OCT) is a new non-contact method of imaging the anterior segment that provides visualisation and measurement of the anterior chamber angle (ACA), and a means to evaluate eyes at risk of primary angle-closure glaucoma.6–12 Traditionally, for quantitative measurements of ACA width, angle opening distance (AOD), trabecular–iris space area (TISA) and trabecular–iris angle are measurements made relative to an anatomical landmark, the scleral spur (SS), which lies approximately 500 μm posterior to the trabecular meshwork (TM). Although AS-OCT provides an objective and quantitative approach to evaluate the ACA, the localisation of SS is sometimes difficult.13 14 Sakata et al reported that the SS location could not be detected in approximately 30% of ACA quadrants, particularly in the superior and inferior quadrants.13 Therefore, new parameters that are independent of SS marking are needed to better characterise ACA width.
Recently, the development of spectral-domain (or high definition) OCT allows a faster scan speed and higher-resolution imaging of ACA than conventional time-domain AS-OCT, which results in the ability to visualise more detailed structures including the termination of the Descemet membrane (Schwalbe line (SL)) and the TM, in addition to SS.15–17 In this study, we propose and evaluate novel quantitative parameters based on a new anatomical landmark, the SL, for assessing ACA width with a spectral-domain OCT device (Cirrus High Definition (HD-OCT), Carl Zeiss Meditec, Dublin, California).
Materials and methods
In this prospective hospital-based study, we recruited consecutive subjects from the glaucoma clinics at the Singapore National Eye Centre. Written informed consent was obtained from all participants, and the study had the approval of the hospital's institutional review board and adhered to the tenets of the Declaration of Helsinki.
All subjects included underwent a full ophthalmic examination, including visual acuity, slit-lamp biomicroscopy, Goldmann applanation tonometry and dark-room gonioscopy. Subjects with history of previous intraocular surgery or penetrating trauma or pigment dispersion syndrome or any cornea opacities or abnormalities that precluded ACA imaging were excluded. However, those who had previously undergone laser iridotomy were not excluded.
Gonioscopy was performed in the dark in all cases by a single examiner with glaucoma fellowship training (CLH), who was masked to imaging findings. A 1 mm light beam was reduced to a narrow slit, and the beam was offset vertically for assessing superior and inferior angles, and offset horizontally for nasal and temporal angles. Static gonioscopy was performed using a Goldmann two-mirror lens at high magnification (×16), with the eye in the primary position of gaze. Care was taken to avoid light falling on the pupil and to avoid accidental indentation during examination. The angle was graded using the modified Shaffer grading system; a quadrant was defined as closed if the posterior TM was not visible on gonioscopy.18 Dynamic (indentation) gonioscopy was performed to establish the presence or absence of peripheral anterior synechiae using a Sussman four-mirror gonioscope. We excluded the subjects with evidence of peripheral anterior synechiae on indentation.
Anterior chamber angle imaging
The Cirrus HD-OCT model 4000 (software version 3.0, Carl Zeiss Meditec) images the anterior segment of the eye by employing spectral-domain imaging technology. The imaging principle is based on low-coherence interferometry using an 840 nm superluminescent light-emitting diode as the light source. The acquisition rate of the Cirrus HD-OCT is 27 000 A-scans/s, which is about 70 times faster than conventional (time-domain) OCT. The transverse and axial resolutions for posterior eye are 15 and 5 μm, respectively. In this study, ACA imaging was performed using ‘five-line raster’ protocol (scan length 6 mm; 4096 A-scans). The scan length was magnified from 6 mm to 7.2 mm as a 60-dioptre aspheric lens (Volk Optical, Mentor, Ohio) was mounted over the Cirrus imaging aperture, with the same distance from Volk lens to the instrument's entrance pupil as a routine fundus examination.15 An external fixation light was used to guide the patient's fixation to the side of the instrument during imaging. Only nasal and temporal quadrants from each selected eye were imaged owing to technical difficulties in scan acquisition of the superior and inferior quadrants (ie, two Cirrus ACA images for each eye). We selected only the scans centred on the corneal limbus from the ‘five-line raster’ series for analysis. Images of poor quality, such as those due to media opacities, motion artefact or images that were out of focus, or which had a poor contrast were excluded.
Quantitative assessment of anterior chamber angle
We proposed two new parameters independent of SS for assessing ACA quantitatively: Schwalbe line-angle opening distance (SL-AOD) and Schwalbe line-trabecular–iris space area (SL-TISA). The SL-AOD was defined as the distance from the SL (the termination of the Descemet membrane) to the anterior iris surface, perpendicular to the corneal endothelial surface. The SL-TISA was defined as an area bounded anteriorly by the SL-AOD, posteriorly by a line drawn along the trabecular meshwork at 500 μm from the SL, perpendicular to the plane of the inner sclera wall to the opposing iris, superiorly by the inner corneoscleral wall and inferiorly by the iris surface (figure 1A).
The built-in Cirrus software version 3.0 (used in this study) cannot correct refractive distortion of anterior segment images. Therefore, we wrote a program using MatLab (version 7.0; the Math Works, Natick, Massachusetts) to dewarp the Cirrus anterior segment images for correcting image misalignment.19 We corrected for refraction both at the air–cornea interface (refractive index=1.38) and at the endothelium–aqueous boundary (refractive index=1.32). We then wrote another program using ImageJ software (NIH ImageJ; National Institutes of Health, Bethesda, Maryland) to measure the SL-AOD and SL-TISA automatically after manual location of the SL. This program also allows the measurement of conventional ACA SS parameters (SS-AOD 500, SS-TISA 500, SS-AOD 750 and SS-TISA 750) after manual location of SS. The SS-AOD 500/750 was defined as the distance from the corneal endothelium to the anterior iris surface perpendicular to a line drawn at 500/750 μm from the SS.20 21 The SS-TISA-500/750 was defined as an area bounded anteriorly by the SS-AOD500/750; posteriorly by a line drawn from the SS perpendicular to the plane of the inner scleral wall to the opposing iris; superiorly by the inner corneoscleral wall; and inferiorly by the iris surface.22
Three observers (CYC, CZ and TAT) selected and identified the location of SS and SL together, and at least two observers agreed on the location. We included only scans in which both SS and SL could be identified for further analysis. A subset of 23 images (13 with open-angle and 10 with closed-angle) were randomly selected and measured by two observers (CZ and TAT) independently to determine the interobserver reliability. One observer (TAT) repeated the measurements after 2 weeks to determine the intraobserver reliability.
One eye from each subject was selected randomly for analysis. Statistical analyses were performed using SPSS statistics version 17.0 (SPSS) and R version 2.11.1 (The R Foundation for Statistical Computing). Interobserver and intraobserver reliability was assessed using intraclass correlation coefficients. The associations between SL parameters, SS parameters and gonioscopic grading were evaluated using the Spearman and Pearson correlation coefficients adjusting for quadrant location (ie, nasal or temporal). We then calculated the z values to test whether two correlations (SL parameters vs SS parameters) have different strengths with gonioscopic grading. The ACA width was compared between open and closed angles with an independent-samples t test. Receiver operating characteristic (ROC) curves and area-under-the-ROC curves (AUC) were used to assess the ability of SL and SS parameters to differentiate between open and closed angles after adjusting for quadrant location.
Seventy-three (47 females, 26 males) subjects were recruited; the majority of the subjects were Chinese (n=65, 89%), with smaller numbers of Malays (n=3, 4%) and Indians (n=5, 7%). The mean (SD) age of the subjects was 60.6 (11.4) years. We excluded 29 images (19.9%) owing to poor image quality, leaving 117 HD-OCT images (65 nasal, 52 temporal) for the final analysis. SL was identified in 95% of ACA scans compared with SS, which was identified in only 85% of images (p=0.035). Both SS and SL were identified in 94/117 (80.3%) scans. SL alone was identified in 17/117 (14.5%) scans, and only SS but not SL was identified in 6/117 (5.1%) scans (McNemar test, p=0.035). We included only those scans (n=94) with both SS and SL visible for the ACA width measurement in the following analysis. Figure 2 shows an HD-OCT image in which both SS and SL were identified. In the 94 scans with both visible landmarks, the mean (SD) SL-AOD, SL-TISA and SS-to-SL distance were 0.25 (0.15) mm (ranging from 0 to 0.64 mm), 0.086 (0.054) mm2 (ranging from 0 to 0.242 mm2) and 0.67 (0.13) mm (ranging from 0.39 to 1.23 mm), respectively.
Table 1 shows the interobserver and intraobserver reliability analysis for the measurements of SL and SS parameters using the customised program. The interobserver and intraobserver reliability were high for both SL and SS parameter measurements (intraclass correlations ranging from 0.98 to 1.00). SL-AOD and SL-TISA were significantly correlated with SS-AOD and TISA parameters (all correlation coefficients ≥0.85, all p values <0.001) and gonioscopic grading (all Spearman ρ correlation coefficients ≥0.69, all p values <0.001) (table 2).
Using gonioscopic grading as the reference, 58 scans with open angles and 36 scans with closed angles were evaluated. The values of SL-AOD, SL-TISA, SS-AOD 500, SS-TISA 500, SS-AOD 750 and SS-TISA 750 were significantly smaller in quadrants with closed angles (all p values <0.001) (table 3). There were no significant differences in SS-to-SL distance between open and closed angles (p=0.498). The discriminatory ability of SL and SS parameters in identifying open and closed angles was comparable (figure 3; all AUCs between 0.85 and 0.90, all p values >0.05). Table 4 shows the correlations of SL and SS parameters with gonioscopic grading in open and closed angles. SL-AOD (r=0.441, p=0.005) and SL-TISA (r=0.433, p=0.006) were significantly correlated with gonioscopic grading in closed angles. There were no significant differences between the correlations of SL parameters and SS parameters with gonioscopic grading (p values ranging from 0.402 to 0.753; data not shown).
In this study, we designed and evaluated two new parameters based on the SL as a new anatomical landmark, independent of SS location, to quantify ACA width. We showed that these new SL parameters (SL-AOD and SL-TISA) have a high reproducibility and showed a good correlation with conventional SS parameters (SS-AOD and SS-TISA) and standard gonioscopic grading.
Early detection and appropriate management/treatment for narrow angles are important strategies to reduce the risk of developing angle closure and angle-closure glaucoma. Quantitative measurements of ACA can also be used to screen and monitor progressive narrowing of the angle and may also provide a better understanding of the inter-relationship between anatomical structures and angle closure. Traditionally, AOD, TIA, angle recess area and TISA have been used to assess ACA quantitatively with ultrasound biomicroscopy and time-domain AS-OCT.20–23 These measurements of ACA width are made with reference to the SS location, anatomically representing the junction of the ciliary muscle fibres and the anchoring region of the TM base. However, the localisation of SS is difficult and often unidentifiable in time-domain AS-OCT machines.13 14 Higher-resolution imaging of ACA allowing visualisation of SL is now available with HD-OCT. We showed that SL can be identified in 95% of scans, thus allowing SL parameters to be measured more often, as compared with SS, which can be identified in only 85% of eyes. This finding was similar to that in a previous study.15 The visibility of SL was better than that of SS, suggesting that SL may be a more useful anatomical landmark for quantitative measurement of the ACA using Cirrus HD-OCT.
Although the TM can also be identified with HD-OCT, the visibility is relatively low (about 60%).15 Therefore, in this study, only SL was used as a landmark to assess ACA width objectively and quantitatively. SS parameters are mostly based on a 500 μm measurement distance, which is the approximate length of the TM based on UBM studies.20–22 However, it is not known whether measurements made by UBM imaging are applicable to HD-OCT imaging. The location at 500 μm may not truly reflect ‘iris-trabecular contact’ for an individual eye. In this study, the SS-to-SL distance was reported, which estimates the length of TM. Our results found that the mean (SD) SS-to-SL distance was 0.67 (0.13) mm, and the range varied from 0.39 to 1.02 mm. In addition, this variability between SS and SL implies that SS parameters may not be consistently measuring the distance between iris and angle at the level of the TM. We therefore believe that SL-AOD, measured at SL, which is situated anterior to the TM, is more appropriate to assess and quantify ‘iris-trabecular narrowing.’
We should point out that SL-AOD may not truly reflect the ACA in some cases, as iris curvature is also important in determining angle closure. Figure 1A,B illustrate two examples with same value of SL-AOD but the angle in figure 1B is more occludable than the angle in figure 1A, thus demonstrating the limitation of SL-AOD. Therefore, we also propose SL-TISA measured at 500 μm from the SL as another SL parameter for ACA width because it takes into account the iris surface irregularity and excludes the non-filtering region behind the SS.22 Our results showed that both SL-AOD and SL-TISA were effective for ACA assessment. The correlations of SL-AOD and SL-TISA with SS parameters and gonioscopic grading were both strong. Furthermore, the discriminatory ability of SL-AOD and SL-TISA in identifying open and closed angles was comparable with that of SS parameters. However, we arbitrarily chose 500 μm from the SL to quantify SL-TISA in this study; further research is required to determine the most suitable distance for the measurement of SL-TISA to reflect the width of the ACA.
Importantly, in analyses stratified by open- and closed-angle groups, we found that SL parameters still showed strong correlations with gonioscopic grading in closed angles. As the SL is situated anterior to TM, and SS is situated posterior to the TM, SS parameters may be measured as ‘zero’ for eyes with partial iris–trabecular meshwork contact, even though the angle is not completely closed, as illustrated in figure 4. Our findings showed that SL parameters may be more representative of angle width and may better reflect standard gonioscopic grading in making assessments for closed angles.
A limitation of this study was that an aspherical lens was mounted externally on the Cirrus HD-OCT for the angle imaging. The Cirrus HD-OCT has an incorporated anterior segment imaging modality on the latest model (version 4.0), and it is not known if the image resolution and quality are the same. However, this was not yet available at the time of this study. Second, manual localisation of SL may still introduce variability in the measurement of SL parameters. Development of a more advanced algorithm with fully automated measurement to minimise the variability is needed for reliable and accurate assessment of angle narrowing using the SL as a landmark. Third, the visibility of SL and SS in the superior and inferior quadrants by HD-OCT has not yet been assessed; it is possible that the pattern of visibility is different in these quadrants. Fourth, although the examiner attempted to acquire ACA image with HD-OCT and perform gonioscopy at the 3 and 9 o'clock positions, the exact location may have varied. Moreover, the patient's fixation has not yet been standardised in the previous (or current) version of Cirrus, and an external fixation light was used during angle imaging in this study. It is unclear whether the visibilities of the landmarks are affected because of this limitation. Other limitations of this study include a relatively small sample size and also a lack of assessment of vertical meridians.
In summary, using HD-OCT, the SL could be identified in 95% of ACA scans and appeared to represent a better anatomical landmark for assessing ACA. Our results suggest that the newly proposed SL-based parameters (SL-AOD and SL-TISA) may be potentially more useful in quantifying ACA width and for assessing risk of angle closure and angle-closure glaucoma. Further studies are required before we can utilise this parameter in clinical management.
Funding Grants from the National Medical Research Council and the National Research Foundation, Singapore.
Competing interests TYW and TA have received grant funding from Carl Zeiss Meditec, but this was not related to the subject of this study. TA has received travel support and honoraria from Carl Zeiss Meditec.
Patient consent Obtained.
Ethics approval Ethics approval was provided by the Singapore Eye Research Institute.
Provenance and peer review Not commissioned; externally peer reviewed.