Relationship Between Iris Surface Features and Angle Width in Asian Eyes

Abstract

Purpose.: To examine the associations between iris surface features with anterior chamber angle width in Asian eyes.

Methods.: In this prospective cross-sectional study, we recruited 600 subjects from a large population-based study, the Singapore Epidemiology of Eye Diseases (SEED) study. We obtained standardized digital slit-lamp iris photographs and graded the iris crypts (by number and size), furrows (by number and circumferential extent), and color (higher grade denoting darker iris). Vertical and horizontal cross-sections of anterior chamber were imaged using anterior segment optical coherence tomography. Angle opening distance (AOD), angle recess area (ARA), and trabecular-iris space area (TISA) were measured using customized software. Associations of the angle width with the iris surface features in the subject's right eyes were assessed using linear regression analysis.

Results.: A total of 464 eyes of the 464 subjects (mean age: 57.5 ± 8.6 years) had complete and gradable data for crypts and color, and 423 eyes had gradable data for furrows. After adjustment for age, sex, ethnicity, pupil size, and corneal arcus, higher crypt grade was independently associated with wider AOD750 (β [change in angle width per grade higher] = 0.018, P = 0.023), ARA750 (β = 0.022, P = 0.049), and TISA750 (β = 0.011, P = 0.019), and darker iris was associated narrower ARA750 (β = −0.025, P = 0.044) and TISA750 (β = −0.013, P = 0.011).

Conclusions.: Iris surface features, assessed and measured from slit-lamp photographs, correlated well with anterior chamber angle width; irises with more crypts and lighter color were associated with wider angle. These findings may provide another imaging modality to assess angle closure risk based on iris surface features.

Introduction

The iris plays an important role in angle closure disease. Many features of the iris, including its geometric measurements and dynamic features, have been linked to angle closure.^1–7 Irises that are thicker, have larger cross-sectional area, and more convexity are associated with higher risk of angle closure.^5,6,8 Furthermore, greater iris volume expansion on pupil dilation also has been linked to angle closure.¹ However, the assessment of some of these iris parameters (e.g., thickness, cross-sectional area) requires technical expertise and costly instruments, such as anterior segment optical coherence tomography (AS-OCT) or ultrasound biomicroscopy. 

In contrast, the surface of the iris has distinctive clinical features, such as crypts, furrows, and color, which can be more readily captured and assessed by standard slit-lamp examination or slit-lamp photography.⁹ These features may be novel markers of angle closure risk. The current assessment methods of iris surface features^10,11 were developed primarily for lighter Caucasian eyes. However, because Asian individuals are at a higher risk of developing primary angle closure glaucoma,¹² measuring and profiling the iris of Asian eyes is more important and clinically useful. 

We recently developed standardized methods to measure iris crypts and furrows in Asian eyes, and showed that these features are correlated with iris thickness.⁹ In this study, we determined the association between surface features of the iris and anterior chamber angle width as measured using AS-OCT. 

Methods

Subjects of this study were enrolled from the Singapore Epidemiology of Eye Diseases (SEED) study, which is a population-based cohort study of eye diseases in Chinese, Malay, and Indian populations aged 40 to 80 years in Singapore.^13–15 The baseline examination was conducted between 2004 and 2011 and a follow-up examination has been conducted since January 2011, 6 years after the baseline examination. For this study, we prospectively recruited 600 consecutive subjects from SEED participants who attended the follow-up examination between August 2011 and April 2013. We excluded eyes with previous ocular surgery (including cataract surgery and filtering surgery) or with previous iris laser treatment, or under IOP-lowering medication, as these conditions may have changed the iris color or morphology in some eyes. Furthermore, eyes with corneal opacity or marked corneal arcus covering 50% or more of the iris area were excluded, as these conditions may affect iris grading. 

Participants recruited in the current study underwent standardized ophthalmic examination according to the study protocol,^13,15 including slit-lamp examination by trained ophthalmologists and AS-OCT imaging. In addition, standardized slit-lamp iris photography was done according to a standardized protocol developed specifically for this study purpose.⁹ Written informed consent was obtained from all participants after explanation of the nature of the study. The study adhered to the tenets of the Declaration of Helsinki, and ethics committee approval was obtained from the Singapore Eye Research Institute Institutional Review Board. 

Iris Photography and Grading

A slit-lamp digital camera (DC3; Topcon Corporation, Tokyo, Japan) was used to take color photographs of the iris at ×16 magnification. Photographs were taken in a dark room (20 lux). The slit-lamp light beam was kept at full width (>20 mm) and height (14 mm) at 30% of the maximum brightness, and angled at 45° temporally. No flash was used. For grading purposes, photographs were viewed on a 1366 × 768/60-Hz resolution screen, using the viewing software ACDSee Photo Manager Version 11.0 (ACD Systems, Seattle, WA, USA). 

We developed a grading scheme for iris crypts, contraction furrows, and iris color as described previously (Supplementary Material).⁹ In brief, for grading of crypts, irises were given an integer grade between 1 and 5 based on the number and size of crypts present, as follows: grade 1, no crypts; grade 2, one to three crypts; grade 3, at least four crypts 1 mm or smaller in diameter; grade 4, at least four crypts 1 mm or larger in diameter; and grade 5, numerous crypts 1 mm or larger in diameter, covering nearly the entire iris. Furrows were graded based on their number and circumferential extent: grade 1, no furrow; grade 2, five or fewer furrows present extending 180° or less); and grade 3, five or more furrows present extending 180° or more). Iris color was graded from 1 to 5 based on the overall color of the iris, in comparison with the reference photographs. If an iris was deemed to fall between two consecutive grades, the higher grade was assigned. As furrows are mostly peripherally located, the grading of furrows was challenging in eyes with extensive corneal arcus. Therefore, eyes with corneal arcus covering 50% or more of the iris area were excluded from further analysis. We previously reported that this grading system has good intra- and intergrader agreements (0.901–0.925 and 0.718–0.836, respectively) as well as excellent intra- and intervisit repeatability (0.976–1.00 and 0.903–1.00, respectively).⁹ 

Anterior Segment OCT Imaging

Anterior segment OCT imaging (Visante; Carl Zeiss Meditec, Jena, Germany) was performed on all participants under standardized dark conditions (20 lux). We used the Anterior Segment Single-Scan protocol to obtain cross-sectional images of the anterior segment along the horizontal axis (nasal-temporal angles at 0°–180°) and vertical axis (superior-inferior angles at 90°–270°). Images were processed using the customized software, the Zhongshan Angle Assessment Program (Guangzhou, China).¹⁶ Following the user input on the location of the two scleral spurs, the algorithm automatically measures the following angle parameters: angle opening distance (AOD250, AOD500, and AOD750), angle recess area (ARA750), and trabecular-iris space area (TISA500 and TISA750). Angle opening distance 250, AOD500, and AOD750 were the lengths of a line from the anterior iris to the corneal endothelium perpendicular to a line drawn along the trabecular meshwork at 250, 500, and 750 μm from the scleral spur, respectively.¹⁷ Angle recess area 750 was defined as the area bordered by the anterior iris surface, corneal endothelium, and a line perpendicular to the corneal endothelium drawn to the iris surface from a point at 750 μm anterior to the scleral spur.¹⁸ Trabecular-iris space area 500 and TISA750 were the areas modified by not including the area posterior to a line drawn from the scleral spur to the anterior iris perpendicular to the plane of the inner scleral wall. Among the three AOD and two TISA measures, AOD750 and TISA750 have been shown to have the highest diagnostic performance for angle closure, respectively.¹⁹ In addition, the software was also used to measure iris thickness, iris curvature (I-Curv), and pupil size. Iris thickness was measured at 2000 μm from the scleral spur. I-Curv was measured by drawing a perpendicular line from the iris pigment epithelium at the point of greatest iris convexity, to the line extending from the most peripheral to the most central points of iris pigment epithelium.⁸ The average of measured values of the four quadrants (nasal, temporal, superior, and inferior) was used for the analysis. Anterior segment OCT image analysis was done without the knowledge of the iris grading results. 

Statistical Analysis

Statistical analysis was performed using MedCalc Version 12.3 (MedCalc Software, Ostend, Belgium) and SPSS Version 20.0 (SPSS, Inc., Chicago, IL, USA). Data from both eyes were collected but only the right eye was used so as to eliminate intereye correlation issue in analysis. 

The association between iris surface features (independent variables) and angle width (dependent variables) was assessed by linear regression models, adjusted for potential confounders, such as age, sex, ethnicity, pupil size, and corneal arcus. Corneal arcus might make the iris color appear lighter, and might partially or completely block the view of the peripheral iris. In addition to excluding eyes with extensive corneal arcus as mentioned earlier, because 59.1% of our study populations had mild corneal arcus, we included presence of corneal arcus in our multiple regression models as a covariate to account for its potential effect on iris grading. We did not adjust for iris thickness, because iris thickness is likely an intermediate variable on the causal pathway from iris surface texture to angle width and, hence, including it in the regression models may lead to overadjustment, thus obscuring the association between iris surface textures and angle width. 

Results

A total of 600 subjects were recruited for this study. Among them, 136 eyes were excluded for the following reasons: pseudophakic eyes (25 eyes), history of laser peripheral iridotomy (2 eyes), inability to complete AS-OCT imaging (6 eyes), poor-quality AS-OCT images in at least one quadrant (63 eyes), unidentifiable scleral spur in at least one quadrant (23 eyes), software demarcation error (13 eyes), and corneal opacity (4 eyes). Furthermore, 41 eyes were ungradable for furrows because of marked or extensive corneal arcus covering 50% or more of the entire iris area. None of the subjects were under any IOP-lowering medication or had undergone filtering surgery. Table 1 shows that compared with the excluded subjects, the included subjects were younger and had a higher proportion of females, but there was no difference in IOP and vertical cup-to-disc ratio between the two groups. In total, 464 eyes were graded for crypts and color, and 423 eyes were graded for furrows and included in the analysis. 

Table 1

 

Comparison of Demographic and Clinical Characteristics Between Included and Excluded Eyes

Characteristics	Included,n = 464	Excluded,n = 136	P Value*
Age, y, mean (SD)	57.5 (8.6)	65.8 (9.5)	0.002
Sex, female (%)	256 (53.9)	69 (50.7)	0.010
Ethnicity			0.303
Malay	234 (50.4)	71 (52.3)
Indian	163 (35.1)	52 (38.2)
Chinese	67 (14.4)	13 (9.5)
IOP, mm Hg, mean (SD)	14.30 (3.09)	14.61 (3.65)	0.428
Vertical cup-disc ratio, mean (SD)	0.41 (0.13)	0.40 (0.14)	0.262

*  Based on independent sample t-test for continuous variables or χ²/Fisher's exact test for categorical variables.

The mean age of the 464 subjects was 57.45 ± 8.61 years, and of them, 53.9% were women. Ocular anterior segment characteristics of the study population are shown in Table 2, and the distribution of grades of iris surface features in the study population is shown in Figure 1. 

Figure 1

 

Distribution of grades of crypts, furrows, and color in the study population.

Table 2

 

Anterior Segment Characteristics of Study Participants, n = 464

Characteristics	Mean (SD) orn (%)
Corneal arcus, present	274 (59.1)
Pupil size, mm	4.01 (0.75)
Iris thickness, mm	0.44 (0.10)
AOD750, mm	0.40 (0.18)
ARA750, mm2	0.31 (0.19)
TISA750, mm2	0.23 (0.10)
I-Curv, mm	0.25 (0.13)

Table 3 presents the association between the three iris surface features and three different measures of angle width, namely AOD750, ARA750, and TISA750. Increasing crypt grade was significantly associated with wider AOD750 (change in width in millimeters per one grade higher, β = 0.021, P = 0.007) and TISA750 (β = 0.012, P = 0.010; Model 1). After adjusting for age, sex, presence of corneal arcus, ethnicity, and pupil size (Model 2), the association between increasing crypt grade and wider AOD750 (β = 0.018, P = 0.023) and TISA750 (β = 0.011, P = 0.019) remained significant, and the association between increasing crypt grade and larger ARA750 (β = 0.022, P = 0.049) became marginally significant. Increasing furrow grade did not show significant association with AOD750, ARA750 and TISA750 before or after adjustment of the aforementioned potential confounders (all P > 0.05). Iris color was negatively associated with all the angle width measures; darker color was associated with narrower AOD750 (β = −0.017, P = 0.048), ARA750 (β = −0.035, P = 0.005), and TISA750 (β = −0.015, P = 0.003). After adjustment for age, sex, ethnicity, corneal arcus, and pupil size, darker color was still significantly associated with smaller ARA750 (β = −0.025, P = 0.044) and TISA750 (β = −0.013, P = 0.011). Figure 2 further illustrates the trend observed between iris surface features and TISA750, independent of age, sex, ethnicity, pupil size, and corneal arcus, showing the increase in TISA750 with increasing crypt grade, and the decrease in TISA750 with increasing color grade. 

Figure 2

 

Relationship between iris features and TISA750. Increasing crypt grade and decreasing color grade were associated with wider TISA750, whereas furrow grade did not show any trend with TISA750. Data shown are after adjustment for age, sex, ethnicity, pupil size, and corneal arcus.

Table 3

 

Changes in Angle Width Per Grade Increase in Iris Surface Features

	Model 1*		Model 2†
	β (95% CI)	P for Trend	β (95% CI)	P for Trend
Crypt, n = 464
AOD750, mm	0.021 (0.006–0.037)	0.007	0.018 (0.002–0.033)	0.023
ARA750, mm2	0.021 (−0.001–0.044)	0.066	0.022 (0.001–0.045)	0.049
TISA750, mm2	0.012 (0.003–0.021)	0.010	0.011 (0.002–0.020)	0.019
Furrow, n = 423
AOD750, mm	0.016 (−0.007–0.039)	0.177	0.004 (−0.019–0.028)	0.737
ARA750, mm2	−0.005 (−0.039–0.029)	0.778	−0.007 (−0.042–0.028)	0.689
TISA750, mm2	0.004 (−0.010–0.018)	0.564	−0.001 (−0.015–0.013)	0.863
Color, n = 464
AOD750, mm	−0.017 (−0.034 to −0.001)	0.048	−0.015 (−0.032–0.002)	0.079
ARA750, mm2	−0.035 (−0.058 to −0.011)	0.005	−0.025 (−0.049 to −0.001)	0.044
TISA750, mm2	−0.015 (−0.025 to −0.005)	0.003	−0.013 (−0.022 to −0.003)	0.011

 β, change in AOD (in mm) or area (in mm²) for one grade higher in iris surface features; CI, confidence interval.

*  Model 1: unadjusted models.

†  Model 2: adjusted for age, sex, ethnicity, pupil size, and corneal arcus.

Table 4 shows the relationship between iris surface features and three additional measures of angle width, namely AOD250, AOD500, and TISA500. Similar trends were observed; higher crypt grade and lower color grade were associated with wider angle, whereas furrow grade was not associated with angle width. 

Table 4

 

Changes in Additional Measures of Angle Width (AOD at 250 μm and 500 μm, and TISA at 500 μm) per Grade Increase in Iris Surface Features

	Model 1*		Model 2†
	β (95% CI)	P for Trend	β (95% CI)	P for Trend
Crypt, n = 464
AOD250	0.011 (−0.001–0.021)	0.055	0.010 (−0.001–0.021)	0.065
AOD500	0.012 (−0.001–0.024)	0.078	0.010 (−0.003–0.022)	0.130
TISA500	0.012 (0.003–0.021)	0.010	0.011 (0.002–0.020)	0.019
Furrow, n = 423
AOD250	0.003 (−0.001–0.024)	0.078	0.010 (−0.018–0.016)	0.130
AOD500	0.010 (−0.008–0.029)	0.276	0.002 (−0.017–0.022)	0.805
TISA500	0.016 (−0.007–0.039)	0.177	0.004 (−0.019–0.028)	0.737
Color, n = 464
AOD250	−0.018 (−0.029–0.024)	0.078	−0.015 (−0.027 to −0.003)	0.011
AOD500	−0.017 (−0.030 to −0.003)	0.016	−0.015 (−0.028 to −0.001)	0.037
TISA500	−0.017 (−0.034 to −0.001)	0.048	−0.015 (−0.032–0.002)	0.079

*  Model 1: unadjusted models.

†  Model 2: adjusted for age, sex, ethnicity, pupil size, and corneal arcus.

Discussion

The surface features of the iris are easy to observe clinically and can be captured with slit-lamp photography. Previously we developed a grading system tailored for Asian eyes to quantify iris surface features and showed that more iris crypts were associated with thinner iris thickness, whereas irises with more furrows tended to be thicker.⁹ In this study, we further examined the associations of these surface features with angle width, and showed that iris crypt and color were significantly associated with angle width. We showed that eyes with larger and more iris crypts had wider angle, and eyes with darker irises had narrower angle. 

Iris crypts are regions of iris stroma hypoplasia or atrophy.¹⁰ On slit-lamp examination, they appear as wide pits on the iris surface, often revealing less dense stroma at the base of the pits. Our results showed that higher crypt grade was associated with wider angle. We speculated that the presence of more and wider crypts may indicate a more porous iris stroma, which permits better flow of aqueous humor across the iris. Such flow may help maintain a low pressure gradient between the posterior chamber and anterior chamber, and in turn reduce the likelihood of the iris bowing forward and narrowing the anterior chamber angle. In addition, we also found that eyes with darker irises had narrower angle. This might be because darker irises have higher melanocytes and melanin content,^10,20 which makes the iris stroma thicker and thus less permeable to aqueous humor. However, further adjustment for iris curvature in our analysis had no effects on the association results (data not shown). This indicates that there may be another mechanism underlying the association between iris crypt or color and angle width that is independent of iris bowing. 

We did not observe any statistically significant association between iris furrows and angle width. Furrows indicate sites on the iris that are frequently folded when the pupil dilates.¹⁰ They appear as protuberant lines on the iris periphery. It could be expected that such protrusions would contribute to the narrowing of the angle; however, this association was not observed. The presence of corneal arcus may have hindered an accurate assessment of the peripherally located furrows. In addition, the exclusion of 41 eyes in which the furrows were ungradable may have reduced our power to detect the association between furrows and angle width. 

Slit-lamp examination of iris surface features is potentially advantageous over the use of AS-OCT. Slit-lamp photography is a simpler method to assess the entire iris globally, permitting the assessment of overall angle width, which is beneficial over the use of the cross-sectional images taken by AS-OCT. Although AS-OCT gives a more direct and precise measurement of angle width, it allows only point measurements of angle width from several cross-sectional images, thus may not represent the whole angle width. Current emerging technology, such as swept-source OCT, allows cross-sectional imaging of the whole 360° view of the anterior chamber, which may allow more accurate measurement of angle width.²¹ However, such instruments are costly and require technical expertise to operate, as opposed to the slit-lamp biomicroscope. 

There are limitations to our study. First, the subjective grading of iris surface features may have introduced grading variability and measurement bias. Nevertheless, the inter- and intragrading agreement was high for all three features in our study.⁹ In addition, the graders were masked from a subject's clinical characteristics while performing iris grading. A more objective, automated method of quantifying iris surface features could be developed to achieve even more precise and reliable measurement of these features. Second, our results may not be directly applied to populations of European descent, whose eyes have much higher crypt frequency^10,22 and lighter color than Asian eyes.^11,23 

In conclusion, iris surface features, particularly iris crypts and color, can provide information on angle width status. We showed that irises with more crypts and lighter color are associated with wider angle. Further studies involving angle closure patients would be warranted to directly assess whether these iris surface features could provide additional predictive power for the risk of angle closure, above and beyond the known risk factors such as shallow anterior chamber depth and shorter axial length. 

Acknowledgments

Supported by grants from the National Medical Research Council, Singapore (NMRC R607/28/2008 and NMRC/NIG/1069/2012). The funding organization had no role in the design or conduct of this research. 

Disclosure: E. Sidhartha, None; M.E. Nongpiur, None; C.Y. Cheung, None; M. He, None; T.Y. Wong, None; T. Aung, None; C.-Y. Cheng, None