Long Anterior Zonules and Angle Closure Disease

Précis: Angle closure disease was present in 59.3% of eyes with long anterior zonules (LAZ). The cause is multifactorial including a thick anteriorly positioned crystalline lens, shorter axial length, and increased lens thickness to axial length factor. Purpose: To study the profile of eyes with LAZ presenting in a glaucoma clinic in a tertiary eye care centre and understand the pathogenesis of angle closure disease in these eyes. Methods: This was a retrospective cross-sectional study. All patients with LAZ seen from January 2014 to December 2018 were included. Demographic and clinical characteristics were noted. LAZ eyes (177 eyes of 177 patients) were compared with an equal number of age and sex-matched controls. LAZ was defined as radially oriented zonular fibers (both pigmented and nonpigmented), extending central to the normal zonular termination zone on the anterior lens surface >1 mm beyond their usual insertion of 1.42±0.24 mm from the lens equator onto the mid peripheral zone or central to it, as seen on slit-lamp examination, following pupillary dilation by a single examiner. Glaucoma was defined according to the International Society for Geographical and Epidemiological Ophthalmology classification. The following biometric parameters were compared: anterior chamber depth (ACD), axial length (AXL), lens thickness (LT), lens position (LP=ACD+0.5×LT), relative lens position (RLP=LP/AXL); lens thickness to axial length factor (LAF=(LT/AXL)×10). LAZ eyes without angle closure disease were also compared with controls. Results: Mean age of patients with LAZ was 64.8±8.1 years. Of these, 63.3% were females. Angle closure disease was present in 59.3% (105/177) patients. Majority of these eyes were primary angle closure suspects (PACS) (53.3%, n=56). Significant differences were found between LAZ eyes and controls for LT (4.8±0.38 mm vs. 4.49±0.40 mm, P<0.0001), ACD (2.68±0.39 mm vs. 3.0±0.32 mm, P<0.0001), AXL (22.37±0.79 mm vs. 22.94±1.1, P<0.0001), LAF (2.14±0.19 vs. 1.96±0.21, P<0.0001), and LP (5.07±0.37 vs. 5.3±0.25, P<0.0001). Conclusions: Angle closure was present in more than half the eyes with LAZ. Majority of these eyes were PACS or had primary angle closure. LAZ eyes had a thicker lens, shallow AC, a shorter axial length and an increased LAF as compared with age and sex matched normal controls. The presence of LAZ may be an indicator of increased risk for angle closure.

L ong anterior zonules (LAZ) is the presence of abnormally long zonules on the anterior capsule of the crystalline lens of the eye leading to a reduced zonule free zone (Fig. 1A). Sometimes they may appear as brown radial lines on the anterior lens surface when pigment gets entrapped on these zonules and have been referred to as pigmented lens striae in earlier literature (Fig. 1B). 1-3 A Krukenberg spindle and dense trabecular pigmentation may also be present owing to the pigment dispersion (Figs. 1C, D). LAZ has been reported in African American, European, and other races and does not seem to have a racial predilection. [1][2][3][4][5] A prevalence of 1.7% has been observed in a clinic based study among African Americans. 5 Two phenotypes have been reported in the literature. One variant is where the LAZ are associated with late onset retinal and macular degeneration (LORMD). The second phenotype has been reported in females mostly aged over 50 years, hyperopic and without retinal degeneration. The patients with LORMD have a serine 163 arginine (S163R) mutation in the C1q tumor necrosis factorrelated protein 5 gene (C1QTNF5/CTRP5), 6,7 whereas the second phenotype has no known mutation.
There have been anecdotal reports in the literature of narrow angles and plateau iris in some eyes with LAZ ( Figs. 2A, B). 8,9 These eyes may also pose a challenge during cataract surgery. 10 In addition it has been suggested that the aqueous outflow might be impaired in some eyes with LAZ. A higher intraocular pressure (IOP) 11 and a shorter axial length has been reported in eyes with LAZ in African Americans. 9 These reports suggest a potential association of LAZ with increased IOP and angle closure disease. Despite this only a few studies have looked at the LAZ phenotype, its associations with glaucoma or angle closure disease. The literature pertaining to a definite association with glaucoma and the pathogenesis of glaucoma in these eyes is limited. We decided to look at the clinical profile of eyes with LAZ which presented at a glaucoma clinic in a tertiary eye care center and compare their biometric parameters with age and sex-matched controls to understand any differences which may contribute to the pathogenesis of angle closure disease in these eyes.

METHODS
Our study is a retrospective case series of eyes with LAZ for the time period January 2014 to December 2018 seen in a single glaucoma clinic at a tertiary eye care institution (Sankara Nethralaya Medical Research Foundation, Chennai, Tamil Nadu, India). All study conduct adhered to the tenets of the Declaration of Helsinki. Institutional review board approval was obtained. LAZ was defined as radially oriented zonular fibers (both pigmented and nonpigmented) extending central to the normal zonular termination zone (usually 1.42 ± 0.24 mm from the lens equator) 11-13 on the anterior lens surface onto the mid peripheral zone or central to it as seen on slit-lamp examination following pupillary dilation by a single examiner (M.K.). Eyes with the presence of LAZ, normal IOP (IOP ≤ 21 mm Hg), normal optic nerve head and open angles (on gonioscopy) were termed as normal LAZ eyes. Glaucoma and angle closure disease was defined according to the International Society for Geographical and Epidemiological Ophthalmology classification. 14 Eyes with nonvisibility of the pigmented trabecular meshwork for 180 degrees or more, IOP below 22 mm Hg, and no peripheral anterior synechiae in the angle and a normal optic nerve head were defined as having primary angle closure suspect (PACS) status. Primary angle closure (PAC) was defined as an eye with PACS status and peripheral anterior synechiae and/or elevated IOP or evidence of irido-trabecular contact (pigment patches over the trabecular meshwork) without glaucomatous damage of the optic nerve head; primary angle closure glaucoma (PACG) was defined as PACS status or PAC with evidence of glaucomatous optic nerve head and/or visual field changes.
All patients as a routine, had undergone an undilated slitlamp examination (including examination of the posterior corneal surface), IOP measurement by Goldmann applanation tonometry, 4-mirror gonioscopy, dilated slit-lamp examination (including examination of the anterior lens surface), biomicroscopic examination of the optic nerve head, and indirect ophthalmoscopy. Applanation tonometry was performed before gonioscopy and pupillary dilation. Gonioscopy was performed using a Posner 4-mirror gonioscope in dim illumination, a shortened slit-lamp beam which did not cross the pupillary margin, and the eye in primary gaze. The angle was graded according to the modified Shaffers grading system. 15,16 Indentation was performed with the gonioscope when angle structures were not visible, and the presence of peripheral anterior synechiae was noted. The clinical presentation of LAZ is quite characteristic on slit-lamp examination (both diffuse and on retro-illumination). Thus, all cases with a clinical documentation of LAZ were included. To address any bias, we included all patients which were examined in a single clinic. Demographics of subjects, IOP by Goldmann applanation tonometry, gonioscopy, and clinical details were noted.
Eyes with available biometry and clinical data were included in the analysis. Of 197 patients, biometry data was available for 177 patients. Biometry was performed using Alcon OcuScan RxP Ophthalmic Ultrasound System. Axial length (AL), anterior chamber depth (ACD), and lens thickness (LT) were noted. Lens position (LP = ACD+0.5×LT), relative lens position (RLP = LP/AXL), lens-axial length factor (LAF = (LT/AL)×10) were calculated. Eyes with LAZ (demographic and biometric parameters) were compared with age and sex matched control eyes. A 1:1 case control matching was done. Normal LAZ eyes were also compared with the controls. One eye of each patient was included for this analysis. In The radial crosssection shows an anterior insertion of ciliary body with absence of the ciliary sulcus. The iris root has a steep rise from its point of insertion with anterior bowing and iridotrabecular meshwork contact suggestive of more than 1 mechanism for angle closure. The central anterior chamber depth is 2.2 mm and the anterior posterior lens thickness is 4.4 mm. The ciliary processes are large with an anterior insertion in all clock hours. B, UBM of an eye with LAZ and pseudo plateau iris due to multiple ciliary body cysts. A moderate sized ciliary body cyst* with crowding of angle structures is seen in the cross-sectional view of the anterior segment. The central iris plane is flat. The ciliary sulcus is absent. The central anterior chamber depth is 1.8 mm and the thickness of the crystalline lens is 4.7 mm. patients with bilateral LAZ, the right eye was included for the analysis. The age and sex matched controls were derived retrospectively from normal eyes (IOP ≤ 21 mm HG, normal optic nerve head) which had undergone cataract surgery and had a confirmed absence of LAZ (confirmed absence of LAZ during cataract surgery). The researcher responsible for selection of the controls (A.G.), was masked to the purpose of the study. The controls were randomly selected from a list of consecutive patients who were treated during the same time period.

Statistical Analysis
The distribution of variables was checked using the Shapiro Wilk test and histograms. Paired Student t test was used to compare the normally distributed paired continuous variables. One way analysis of variance and Kruskal-Wallis tests were used to compare multiple groups depending upon the normality of the data. All tests were 2-sided and a P-value < 0.05 was considered statistically significant. Bonferroni adjustment was done for multiple comparisons. Conditional univariate and multivariate logistic regression was performed. All statistical analyses were performed with commercially available software (Stata version 15; StataCorp, CollegeStation, TX).

RESULTS
Of a total database of 338 eyes (197 patients with LAZ) 177 patients (324 eyes) met the inclusion criteria. Of these 147 patients had bilateral LAZ. Mean age of patients with LAZ was 64.8 ± 8.1 years. Mean age of males with LAZ was 66.6 ± 8.6 years and females was 63.8 ± 7.8 years (P = 0.02). Majority of the patients with LAZ were females (63.3%, n = 112) ( Table 1).

Comparison of LAZ Eyes With Controls
Eyes with LAZ had a significantly shorter axial length (P < 0.0001) and shallow central ACD (P < 0.0001) as compared with controls ( Table 1). The lens was significantly thicker as compared with controls with a higher LAF (P < 0.0001). The LP was more anterior (P < 0.0001). There was no significant difference in RLP.

Comparison of Normal LAZ Eyes With Controls
Normal LAZ eyes had a shorter axial length, increased lens thickness and an increased LAF as compared with the controls (Table 2). However, there was no significant difference in ACD, LP, and RLP.
There was no significant difference in the mean age of LAZ patients with angle closure disease and normal LAZ patients (64.6 ± 8.6 y vs. 64.4 ± 6.7 y; P = 1.00). There was no significant difference in the refractive error between normal LAZ eyes and LAZ eyes with angle closure disease (0.71 ± 0.1.7 D vs. 0.98 ± 1.8 D; P = 1.00). Although IOP  was higher in normal LAZ eyes as compared with controls the difference was not significant (14.7 ± 2.6 mm Hg vs. 14.1 ± 2.6 mm Hg; P = 0.18).
Univariate conditional logistic regression analysis showed a significant association of the presence of LAZ with increased IOP, decreased ACD, decreased AXL, increased LT, and increased LAF. On multivariate analysis we found that, on an average, the odds of having LAZ increased 1.3 times for every 1 mm Hg increase in IOP (P < 0.0001) and 164 times (P < 0.0001) for every unit increase in LAF.

DISCUSSION
In our cohort of patients with LAZ, we found that more than half of eyes with LAZ (59.3%) had angle closure disease. Majority of these eyes (80.9%) were either PACS or had PAC. Eyes with LAZ also had a shorter axial length, thicker and an anteriorly placed crystalline lens as compared with controls. A short axial length, decreased central ACD, thick and anteriorly placed lens, and a high LAF are known risk factors for PACG. George et al 17 reported that eyes with occludable angles had shorter axial length and decreased ACD. The reported mean biometric values ( ± SD) from a normal south Indian population are an axial length of 22.76 ± 0.78 mm, ACD of 3.00 ± 0.30 mm and a lens thickness of 4.31 ± 0.31 mm. 17 This is similar to the biometric parameters in the controls in our study. Roberts et al 9 reported an axial length of 23.1 ± 0.92 mm in eyes with LAZ in an African American cohort as compared with controls (23.79 ± 1.01 mm). This variation may be due to a difference in the ethnicity of their study population. Majority of the patients with LAZ were females in our study. This is similar to the distribution noted by Roberts et al 9 and Moroi et al. 18 Females with LAZ were younger than males.
In our study, normal LAZ eyes also had a shorter axial length, increased LT and consequently an increased LAF as compared with controls. However, there was no significant difference in the RLP and ACD. This may explain the absence of angle closure in these eyes despite a shorter axial length and increased lens thickness. It would be interesting to follow up these eyes to look at any forward shift in lens position with time and the incidence of angle closure disease.
There have been a few studies and case reports describing the presence of LAZ in both open angle glaucoma and angle closure disease. Most of the studies have been conducted in a clinic based population where majority of subjects were either African American or Caucasians. 6,8,18 Roberts et al, 11 in their clinic based study with predominantly African American participants found LAZ eyes to have a higher IOP by 1 mm. The authors performed gonioscopy in eyes with a narrow van Herick grade or IOP > 21 mm Hg. They had excluded eyes which were on IOP lowering medication as well as eyes with occludable angles. 11 Angle closure was present in 8.2% cases in their series.
We found eyes with LAZ to present with a spectrum of phenotype ranging from normal IOP, open angles, ocular hypertension, primary open-angle glaucoma, and angle closure disease (Fig. 3). Majority of our subjects had angle closure disease. The difference in phenotype from other studies could be due to a difference in study populations and ethnicity. In addition, our cases are derived from a predominant glaucoma practice where patients with occludable angles, increased IOP, Krukenberg spindles, and pigment dispersion may be referred resulting in us seeing this part of the LAZ spectrum more often. We found that the mean IOP in LAZ eyes was significantly higher than controls. Although the IOP in normal LAZ eyes was higher than controls, the difference was not statistically significant. However, pachymetry values were not available for all the eyes. The trend in IOP was similar to that reported by Roberts et al. 11 Pachymetry data was not available in their study. 11 However, the authors have reported no significant difference in the mean central corneal thickness between African Americans with LAZ (61 subjects) and controls matched on age, race, and sex in a separate study. 9 Ocular biometry plays a significant role in the pathogenesis of angle closure disease. Other factors include zonular laxity, lens iris diaphragm mobility, curvature of the anterior lens surface (lens vault) and an anteriorly positioned crystalline lens. According to Goldberg's 19 model of accommodation based on reciprocal zonular action, the anterior zonules play a role in accommodation rather than the structural support of the lens. They become lax during ciliary body contraction resulting in an increased lens thickness and convexity. We hypothesize that in the presence of a thicker, anteriorly placed lens (Fig. 4), the iris rubs against the long anteriorly placed zonules especially near the pupillary margin causing the pigment to disperse in the anterior segment (Figs. 1B-D). The dispersed pigment may deposit in the trabecular meshwork (Fig. 1D) causing structural changes leading to increased aqueous outflow resistance. This may contribute to an additional decrease in aqueous outflow in the presence of an occludable angle and a compromised trabecular meshwork consequent to the coexistent irido-trabecular contact. An inherent susceptibility to pigment dispersion or pigment related trabecular damage may also be present. However, histopathologic and laboratory studies may be required to confirm this hypothesis. The concomitant presence of plateau and pseudoplateau iris (Figs. 2A, B) may be an additional contributing factor in the pathogenesis of angle closure and increased IOP. Thus, iridotrabecular apposition can be caused by one or more abnormalities at different anatomic sites. LAZ eyes with occludable angles and patchy pigmentation were labeled as PAC. However, we must acknowledge that LAZ eyes are also likely to have patchy pigmentation due associated pigment dispersion. We can apply the sufficient cause model in eyes with LAZ to explain the pathogenesis of increased IOP, angle closure disease and glaucoma. This model has earlier been used to hypothesize the pathogenesis of angle closure disease. 20 We propose a model where a combination of LAZ, pupillary block, thick crystalline lens, anteriorly positioned lens, plateau iris, pseudo plateau iris, pigment dispersion, unknown genetic and environmental factors may cause an increase in IOP, angle closure disease and open angle glaucoma. Each of these factors alone may not be sufficient to cause an increase in the intraocular pressure, glaucoma or angle closure. When present in combination in a varying degree, these factors can result in an increased IOP, angle closure and glaucoma (open angle or angle closure). Different combinations of these factors may be responsible for the spectrum of phenotype seen in eyes with LAZ.
The presence of a shallow anterior chamber, occludable angle, Krukenberg spindle or high IOP during a routine examination may be responsible for the early referral of these patients to a glaucoma clinic. Majority of the eyes with LAZ in our study were PACS or had PAC. It may be interesting to follow up these eyes for progression of the angle closure disease. Primary angle closure disease is a widely used term for the angle closure spectrum (including the American Academy of Ophthalmology Preferred Practice Pattern). 21 We are reporting the association in an Asian population that has a high risk for angle closure. However, in view of the multiple mechanisms involved in the pathogenesis of glaucoma and angle closure disease in eyes with LAZ, one may argue that the glaucoma associated with these eyes can be termed as a secondary glaucoma rather than primary.
The limitation of our study is the retrospective nature. Moreover, our study population includes patients which may present at a glaucoma clinic and may not be representative of the distribution in a normal population. The patients are representative of a cohort which is more likely to be seen in a predominant glaucoma and cataract practice. They may not represent the entire spectrum of cases seen either in the general population or in a primary or secondary level eye care centre. Although majority of the LAZ subjects in our study population had angle closure disease, our cohort also had patients with LAZ who did not have angle closure disease or glaucoma. A small proportion also had ocular hypertension or primary open-angle glaucoma. This may be due to an increased referral of eyes with angle closure disease. However, the large proportion does point toward an association of LAZ with angle closure disease which needs to be validated in future studies. Although gonioscopy data was available for all patients, anterior segment imaging data were not available for all eyes with angle closure. We acknowledge that the proportion of LAZ subjects with angle closure disease may be higher in our study population owing to the fact that patients were derived from a predominant glaucoma practice and may not reflect the distribution in a general population. However, all patients as a routine underwent slit-lamp examination following pupillary dilation. We also included all eyes with a clinical documentation of LAZ instead of only those with clinical photographs to avoid bias. Ours is a retrospective study and only documented cases of LAZ were included. We cannot rule out any mild LAZ cases which may have not been clinically documented. However, we tried to address this by including all LAZ cases seen in a single clinic. It would be interesting to look at the prevalence of LAZ in a population based study.
The normal zonular insertion and the age-related shift 12,13,22 is unlikely to explain the small zonule free zone seen in LAZ eyes. Ours is a cross-sectional study and it would have been interesting to prospectively see the changes in insertion of zonules, lens thickness and angle of the anterior chamber over a period of time.
Being a retrospective study, we tried to address any selection bias by including patients seen in a single clinic by a single examiner to maintain reliability and consistency. Biometry was advised for all patients detected with LAZ irrespective of the plan for cataract surgery or presence a visually significant cataract, including patients with a clear lens. This was due to our initial experience with these eyes, in which we noted the presence of an occludable angle on gonioscopy or a thick lens intraoperatively. Of all the LAZ patients, 20 did not undergo biometry, hence they were excluded from the analysis. Only 52 underwent cataract surgery. Patients with a clear lens or where the cataract was not visually significant also underwent biometry. Since the lens thickness increases with an increase in age, 23,24 our cases were age and sex matched to controls. Although there is an assumption that the presence of cataract or increased lens density is associated with increased LT, reports in the literature regarding this are contradictory. [24][25][26][27] Some studies have found that LT is independent of lens density in eyes with mild to moderate cataracts after accounting for age. 26 On the other hand, some studies have reported an increase in the LT in eyes with nuclear sclerosis as compared with clear lenses whereas some have reported no difference or a weak correlation. [24][25][26][27] The presence of LAZ has been reported in an older age group in the literature. Roberts et al 4,5 rarely observed LAZ in eyes of subjects younger than 50 years of age in their study population, despite actively searching for them. They hypothesized that this could be due to a lengthening and/or hypertrophy of zonular fibers over time. Reiger 28 also conducted a detailed examination for the presence of LAZ following mydriasis in their hospital based study population in Austria. They reported that subjects with LAZ had a mean age of 67.9 years with the youngest being 50 years of age. They also examined 1108 children and young subjects after pupillary dilation and did not find any evidence of LAZ. They hypothesized that the presence of LAZ was associated with aging. Majority of the patients in our cohort were more than 50 years of age, with a mean age of 64.8 ± 8.1 years. Some degree of cataract is likely to be present in this age group. Fifty-two patients in our cohort underwent cataract surgery and 98.3% (174/177 eyes) had cataract (both visually significant cataract, as well as not visually significant). Thus, the association of LAZ with cataract needs to be investigated by prospective population based studies.
In African American females with LAZ (50 subjects), no significant difference in ocular biometric parameters was observed when LAZ eyes were compared with refractive error and age matched controls. 29 Fifty of the total 92 LAZ subjects had participated in this study. Two patients with LAZ were found to have occludable angles. This could be due to the difference in ethnicity from our cohort. Moreover, the mean age of this cohort was 67.1 years, in whom cataract (nuclear sclerosis) is likely to be present and act as a confounder during matching, due to cataract induced refractive error. It will be interesting to compare the incidence of angle closure in LAZ eyes with controls having a similar refractive error.
Being a retrospective study, we used biometry data from patients who had undergone cataract surgery for our controls. The lens thickness is related to aging. The cases and controls were age matched in our study. Evidence in the literature is unclear regarding the association of nuclear sclerosis and lens thickness. [24][25][26][27] Assuming that nuclear sclerosis results in increased lens thickness as a worst case scenario, we may be underestimating the difference in the lens thickness between LAZ eyes and controls. The lens was significantly thicker as compared with controls in our cohort, with a significantly higher LAF and a more anterior LP. The biometric parameters of our controls were also similar to that reported in the normal south Indian population. 17 Thus, our choice of controls is conservative with a possibility of an underestimation of the difference between lens thickness between the 2 groups. We feel that the relation between LAZ, cataract and angle closure disease is complex and multifactorial. It could be related not only to the increased lens thickness but also the lens position, presence of plateau iris or pseudoplateau iris, the dynamic changes during pupillary movement, pigment dispersion and the sensitivity of the trabecular meshwork to pigment-related damage. Since the anterior zonules play a role in accommodation, 19 the zonules may be part of the pathogenesis of increased lens thickness, even though accommodation decreases with an increase in age. The excursion of the capsule-zonular attachment is still required to produce the small amount of accommodation in the aging eye. 30 The increased lens thickness may be due to multiple mechanisms and complex interactions between the zonules, lens capsule, and ciliary body. In addition, a subset of our patients had open angle glaucoma. All these are avenues for future research. Population based prospective studies may shed further light on this association.
Of the 2 described LAZ phenotypes, our cases seem to belong to the variant without retinal degeneration although 2 patients had a family history of glaucoma. An asymptomatic stage 1 of this condition has also been described where LAZ is the only clinical sign with normal dark adaptation. 31 The phenotype described in our series (LAZ without LORMD) has no known mutation. It would be worth looking at the genetic profile of this cohort.

CONCLUSIONS
The presence of LAZ may be a surrogate marker of angle closure disease or part of a syndrome. More than half the eyes with LAZ in our cohort had angle closure. Majority of these eyes were PACS status or had PAC. LAZ eyes had a thicker lens, shallow AC, a shorter axial length and an increased LAF as compared with age and sex matched normal controls. The mechanism of increased IOP, glaucoma, and angle closure may be multifactorial in eyes with LAZ. These cases need more careful examination and management.