Long-Term Changes in Lamina Cribrosa Curvature Index After Trabeculectomy in Glaucomatous Eyes

Abstract

Purpose: This study aimed to evaluate both short-term and long-term changes in the lamina cribrosa curvature index (LCCI) following trabeculectomy and investigate the factors influencing these changes.

Methods: In this retrospective, observational study, 40 eyes of 40 patients with glaucoma who underwent trabeculectomy and had a follow-up of at least 2 years were included. Optic nerve head area was scanned by using spectral-domain optical coherence tomography before surgery (Pre_OP), within 6 months postoperatively (Post_OP1), and at the last visit (Post_OP2). LCCI values calculated from B-scan images at six different planes (0°, 30°, 60°, 90°, 120°, and 150°) and their mean values were compared. Univariate and multivariate linear regression analyses were used to identify the clinical factors associated with the amount of LCCI changes.

Results: The mean follow-up time was 38.3 ± 16.8 months. At Post_OP1, the mean LCCI decreased from 9.28 ± 2.58 to 7.91 ± 2.57 (P < 0.001), and the mean intraocular pressure decreased from 22.0 ± 7.6 mm Hg to 12.2 ± 3.8 mm Hg (P = 0.001). At Post_OP2, the mean LCCI was maintained at 7.74 ± 2.49 (P = 0.56 when compared to Post_OP1 and P < 0.001 when compared to Pre_OP). The mean intraocular pressure was 12.6 ± 5.4 mm Hg (P = 0.67 when compared to Post_OP1 and P < 0.001 when compared to Pre_OP). Long-term LCCI changes were associated with baseline age (P = 0.04), spherical equivalent (P = 0.02), mean IOP during follow-ups (P = 0.02), and preoperative LCCI (P = 0.04).

Conclusions: Glaucomatous eyes undergoing trabeculectomy demonstrated reductions in the LCCI after a mean follow-up of over 3 years. Greater long-term LCCI reduction was associated with younger age, lower mean IOP during follow-up period, greater spherical equivalent refractive error, and preoperative LCCI.

Glaucoma, the leading cause of irreversible blindness worldwide, encompasses a heterogeneous group of optic neuropathies that are characterized by progressive loss of the retinal ganglion cells (RGCs) and their axons.^1,2 Although the mechanisms of glaucomatous optic neuropathy are not fully understood, intraocular pressure (IOP) is believed to play an important role in its pathophysiologic processes.³ Depending on the susceptibility of an individual optic nerve head (ONH), IOP-related stress and strain can cause mechanical failure of the connective tissue in the ONH area and axonal compromise of the RGCs within the lamina cribrosa (LC).⁴ 

The LC is considered the primary site of glaucomatous optic nerve injury, and anatomical changes in the LC have been reported to be associated with glaucoma in ex vivo studies.^5–11 The development of high-resolution imaging techniques has enabled in vivo measurements of LC morphologic parameters such as focal defects, thickness, depth, and curvature index, which have been found to be correlated with the onset and progression of glaucomatous damage.^12–18 Considering that further disease deterioration after IOP-lowering treatment is hypothesized to be caused by inadequate pressure reduction at the LC level,⁴ evaluation of LC morphologic changes after treatment has the potential to serve as a more accurate measure to determine treatment success. 

A few previous studies have described the changes in LC morphology after medical or surgical IOP reduction. Significant anterior movement has been reported after IOP reduction, mostly with short-term to mid-term follow-ups.^19–28 Notably, Lee et al.²⁹ described the reversal of LC displacement over a mean follow-up time of 27.1 ± 3.3 months. However, the morphologic parameter used in their study, LC depth, is measured using the Bruch's membrane opening (BMO) level as a reference plane, which makes it susceptible to influences caused by changes of choroidal thickness. Choroidal thickness is known to vary among individuals³⁰ and is influenced by aging and physical exercise.^31,32 On the other hand, morphological parameters, such as the lamina cribrosa curvature index (LCCI), that evaluate LC curvature are not affected by choroidal thickness, and they have shown better efficacy for discriminating glaucoma and healthy eyes than LC depth.³³ This study aimed to evaluate the long-term changes in the LCCI following trabeculectomy, and explore factors associated with the degree of both short-term and long-term changes. 

Methods

This study was a subanalysis of the Bern Glaucoma Registry Study, a retrospective cohort study at Inselspital, Bern University Hospital in Bern, Switzerland. Ethical approval for the study was granted by the Ethics Committee of the Canton of Bern (BASEC ID 2022-01046). The study protocol was in accordance with the tenets of the Declaration of Helsinki. Consecutive medical records of patients with glaucoma indicated for surgical intervention who underwent trabeculectomy between 2015 and 2020 were retrospectively reviewed for eligibility. The need for participation consent was waived due to the retrospective nature of this study. 

Study Participants

Patients included in this study were required to be over 18 years old, with a best-corrected visual acuity (BCVA) of 0.5 or better, a spherical refraction of −8.0 to +3.0 diopters (D), and a cylinder correction within ±3.0 D. Additionally, patients had to have been followed up for at least 2 years postoperatively and to have optical coherence tomography (OCT) scans in the ONH area at pre- and postoperative visits. The glaucoma diagnosis was made by experienced glaucoma specialists at Inselspital. Criteria for glaucoma diagnosis included (1) evidence of glaucomatous optic nerve damage, and (2) corresponding visual field defects. Glaucomatous optic nerve damage was defined as neuroretinal rim thinning, optic disc excavation, disc hemorrhage, or retinal nerve fiber layer (RNFL) defects. The inclusion criteria had to be met in at least one eye for each patient. If both eyes met the criteria, the eye that underwent surgery first was included for this analysis. 

Patients were excluded if they met the following criteria: (1) a history of previous intraocular surgery (except for uneventful cataract surgery); (2) co-existing ocular or central nervous system diseases that could affect the visual field or OCT measurements; (3) postoperative complications such as hypotony or disc edema (lasting > 6 weeks); or (4) insufficient good-quality OCT scans. Good-quality OCT scans were defined as follows: (1) OCT images with a quality score (Q) of ≥20; (2) visibility of the anterior surface of the LC ≥ 70%; or (3) ability to identify BM termination points. 

Pre- and Postoperative Examinations

Before surgery, demographic and clinical information for each patient, including age, sex, self-reported history of diabetes mellitus and systemic hypertension, family history of glaucoma, glaucoma subtype, and number of anti-glaucoma medications, was documented. On the day before the operation, all patients underwent a comprehensive ophthalmic examination, including measurements of BCVA (recorded as logMAR values), refractive errors (recorded as spherical equivalent), and slit-lamp examination. The mean defect (MD) of visual field testing (Octopus 900 perimeter; Haag-Streit, Köniz, Switzerland), RNFL thickness, BMO area and BMO width by spectral-domain OCT (SPECTRALIS; Heidelberg Engineering, Heidelberg, Germany), and IOP by Goldmann applanation tonometry (Haag-Streit) were also documented. Data on IOP measurements at each postoperative visit were also documented. Mean IOP was defined as the average of all IOP measurements from the preoperative visit until the last visit. 

Measurements of LCCI

The ONH area was scanned at 1 day before surgery (Pre_OP), within 6 months postoperatively (Post_OP1), and at the last visit (Post_OP2). The LC was evaluated at six different B-scan planes (0°, 30°, 60°, 90°, 120°, and 150°), counterclockwise in the right eye and clockwise in the left eye (Figs. 1A, 1B). The LCCI was measured using a built-in caliper tool in HEYEX 2.3.2 (Heidelberg Engineering). The LCCI was calculated as the ration of the LC curve depth to the LC curve width, multiplied by 100 (Figs. 1C, 1D). Lines (BMO width) connecting two BM termination points were drawn first. Then, the LC curve width was assessed by measuring the width of the lines connecting the two points on the anterior LC surface that met the lines perpendicular to the BMO width lines. The LC curve depth was defined as the maximum depth from the line indicating LC curve width to the anterior surface of the LC. The LCCI was measured by two trained observers (JDU, XS) masked to other clinical information. The mean values of the two measurements were used for analysis. The mean LCCI was calculated as the average value of the LCCI measured at all six planes. LCCI change was defined as the value of postoperative measurements minus baseline measurements. 

Figure 1.

 

Measurements of the LCCI. (A, B) En face image of the optic nerve head (A) with six lines intersecting at the center of the optic disc at intervals of 30° indicating the locations where the measurements were performed (B). (C) B-scan image obtained from the location where the red line was in B. (D) Calculation of the LCCI. The lines connecting two BM termination point were drawn first. Then, the LC curve width was assessed by measuring the width of the line connecting the two points on the anterior LC surface that met the lines perpendicular to the BM opening width lines. The LC curve depth was defined as the maximum depth from the line indicating LC curve width to the anterior surface of the LC. The LCCI was calculated as the ratio of the LC curve depth to the LC curve width, multiplied by 100. LCCW, lamina cribrosa curve width; LCCD, lamina cribrosa curve depth.

Statistical Analysis

Statistical analysis in this study was performed using SPSS Statistics 24 (IBM, Chicago, IL, USA) and Prism 9.5.0 (GraphPad Software, Boston, MA, USA). Numerical data are presented as mean ± standard deviation, and categorical data are presented as frequencies and percentages. The agreement between two observers was assessed using the intraclass correlation coefficient. Pre- and postoperative LCCI and IOP values were compared using the Wilcoxon test. Univariate and multivariate linear regression models were employed to evaluate the association between the clinical characteristics and the degree of LCCI change. Variables with P < 0.10 in the univariate models were included in the stepwise multivariate regression model. P < 0.05 was considered statistically significant. 

Results

Baseline Demographic and Clinical Characteristics

A total of 57 eyes with glaucoma were included in this study. Six eyes were excluded for a history of previous glaucoma surgery (two eyes underwent glaucoma drainage device implantation, three eyes underwent deep sclerotomy, and one eye underwent cyclophotocoagulation). One eye was excluded for postoperative cystoid macular edema. Two eyes were excluded for co-existing ocular or central nervous system diseases (one eye for cavernous arteriovenous fistula and one eye for wet age-related macular degeneration). Eight eyes were excluded for poor-quality OCT images. Forty eyes from 40 patients were qualified for the final analysis. Demographic and clinical characteristics of the included subjects at baseline are summarized in Table 1. 

Table 1.

 

Baseline Demographic and Clinical Characteristics of the Study Population

 

The glaucoma patients included in this study had a mean age of 67.4 ± 9.4 years (range, 52–85) at baseline. Seventeen of them were female and 23 male. At the preoperative visit, three of the patients reported diabetes mellitus and 11 of them reported systemic hypertension. Twenty-three of the included eyes were diagnosed with primary open-angle glaucoma (POAG), 13 with pseudoexfoliative glaucoma, and four with primary angle-closure glaucoma. The baseline mean IOP was 22.0 ± 7.6 mm Hg (range, 11–40), and the mean number of anti-glaucoma medications was 3.3 ± 1.5 (range, 0–6). Their mean logMAR visual acuity was 0.14 ± 0.11 (range, 0.0–0.3), and the mean spherical equivalent was −0.84 ± 0.11 D (range, −7.38 to +2.50). The MDs and mean global RNFL thicknesses at baseline were 10.58 ± 6.77 dB (range, 2.2–24.5) and 58.0 ± 13.5 µm (range, 31–90), respectively. The mean follow-up time of our study cohort was 38.6 ± 16.8 months (range, 24–92). 

LCCI and IOP Changes After Trabeculectomy

The intraclass correlation coefficient between two observers was 0.910 (95% confidence interval [CI], 0.886–0.929, P < 0.001). As shown in Table 2, the mean IOP decreased to 12.2 ± 3.8 mm Hg at 6 months postoperatively (compared to Pre_OP, P = 0.001), the mean reduction was 10.6 ± 9.4 mm Hg (39.1% ± 29.6%). At the last visit, the mean IOP was 12.6 ± 5.4 mm Hg (compared to Pre_OP, P < 0.001), and the mean reduction was 9.7 ± 6.2 mm Hg (41.5% ± 21.1%). 

Table 2.

 

LCCI and IOP at Pre_OP, Post_OP1, and Post_OP2

 

At 6 months postoperatively, the mean LCCI decreased significantly from 9.28 ± 2.58 at baseline to 7.91 ± 2.57 (P < 0.001). At the last visit, the mean LCCI was 7.74 ± 2.49 (compared to Pre_OP, P < 0.001). The mean values of long-term LCCI changes at the last visit were −1.54 ± 1.83 (range, −6.46 to 1.21), with 31 out of 40 eyes demonstrating smaller LCCI values compared to preoperative measurements. 

For the long-term LCCI changes at each plane, there was a significant decrease in LCCI for all six planes compared to preoperative measurements at the last visit (all P ≤ 0.02) (Fig. 2). Changes in the LCCI from preoperative to the last visit at the six different planes ranged from −0.77 ± 2.48 to −2.08 ± 2.48 (Table 2). 

Figure 2.

 

LCCI of the six different planes at preoperative and postoperative visits. (A) Comparison of the LCCI measurements between Pre_OP and Post_OP1. The LCCI was significantly smaller in all planes except 0° (all P ≤ 0.02, except P = 0.06 at 0°). (B) Comparison of LCCI measurements between Pre_OP and Post_OP2. The LCCI was significantly smaller in all planes (all P ≤ 0.02). (C) Comparison of LCCI measurements between Post_OP1 and Post_OP2. There was no significant difference between LCCI measurements at Post_OP1 and Post_OP2.

Factors Associated With Short-Term and Long-Term LCCI Reduction

As demonstrated in Table 3, univariate linear regression showed that baseline age (β = 0.489, P = 0.01) and preoperative LCCI (β = −0.400, P = 0.04) were factors affecting the amount of short-term LCCI changes. Multivariate analysis showed that only baseline age had a statistically significant influence on the short-term LCCI changes (β = 0.489, P = 0.01). 

Table 3.

 

Factors Associated With Short-Term Reduction of the Mean LCCI

 

For long-term LCCI change, univariate analysis suggested that baseline age (β = 0.289, P = 0.07), spherical equivalent (β = −0.289, P = 0.07), IOP reduction in percentage (β = −0.320, P = 0.05), mean IOP during follow-ups (β = 0.356, P = 0.02), and preoperative LCCI (β = −0.402, P = 0.01) were statistically significantly associated with long-term LCCI reduction (Table 4). Multivariate linear regression analysis showed that baseline age (β = 0.293, P = 0.04), spherical equivalent (β = −0.336, P = 0.02), mean IOP (β = 0.317, P = 0.02), and preoperative LCCI (β = −0.299, P = 0.04) were associated with long-term LCCI changes. Scatterplots of the relationships of short-term and long-term changes in LCCI with spherical equivalent and preoperative LCCI are shown in Figure 3. 

Table 4.

 

Factors Associated With Long-Term Reduction of the Mean LCCI

 

Figure 3.

 

Scatterplots showing the relationships of associated factors and short-term and long-term reduction of the LCCI. (A) Relationship of age and short-term reduction of the LCCI. (B) Relationship between age and long-term LCCI reduction. (C) Relationship between baseline spherical equivalent refraction and long-term LCCI reduction. (D) Relationship between baseline LCCI and long-term LCCI reduction. (E) Relationship between mean IOP during follow-up and long-term LCCI reduction. SE, spherical equivalent.

Representative Case

A representative case of both the short-term and long-term changes in LC curvature after trabeculectomy is shown in Figure 4. 

Figure 4.

 

Representative case of short-term and long-term reductions of the LCCI after trabeculectomy: (A) En face image of the ONH. (B–D) B-scan images of a 61-year-old male patient who underwent trabeculectomy: preoperative visit (B), 3 months postoperatively (C), and 24 months postoperatively (D). IOP decreased from 20 mm Hg (B) to 10 mm Hg (C) 3 months postoperatively. The IOP measurement at the 24-month postoperative visit was 12 mm Hg (D).

Discussion

In this study, we evaluated both short-term and long-term LC curvature changes in glaucomatous eyes after trabeculectomy. We found that LC curvature decreased significantly within 6 months postoperatively, which remained statistically significant over a mean follow-up of 3 years. A greater degree of short-term reduction in LC curvature after trabeculectomy was associated with younger age, and greater long-term LCCI reductions were associated with younger age, greater spherical equivalent refractive error, larger preoperative LCCI, and lower mean IOP during the follow-up period. To our best knowledge, this study is the first to report a reduction in LC curvature following trabeculectomy with a long-term follow-up of more than 3 years. 

With the recent development of OCT imaging technology, more studies have been focusing on evaluating the deep ONH structures, including LC morphology in glaucomatous eyes. Focal LC defects were reported to be associated with the location of both structural and functional glaucomatous damage,^34–37 and LC thickness was reported to be thinner in glaucomatous eyes and to vary according to disease stages.^14,38–40 However, due to the limitation of image quality in the current OCT models, visibility of the posterior LC surface is not easily achieved in the clinical setting. Previous studies have also reported greater LC depth in glaucomatous eyes, and LC depth was found to be associated with the rates of glaucomatous deterioration.^15,16 LC depth is usually defined as the distance between the anterior surface of the LC and a reference plane at BMO level; therefore, LC depth could be affected by the thickness of the choroid that lies between BM and the sclera.⁴¹ Another parameter to describe the posterior movement of LC is its curvature, usually quantified by the LCCI, which is independent of choroidal thickness. The LCCI has been reported to be greater in glaucomatous eyes and associated with faster disease progression.^33,42,43 

Several studies have evaluated the reversal of LC posterior bowing following IOP-lowering treatment, with most studies focusing on changes in LC depth.^{19,20,22–25,29} Lee et al.²¹ published the first study evaluating the LCCI in horizontal ONH cross-sections in POAG eyes that underwent trabeculectomy. They found a reduction in LC curvature 6 months postoperatively. In our current study, we similarly observed a decrease in the LCCI within 6 months after trabeculectomy. Furthermore, we evaluated the LCCI using six radial ONH scans spaced 30° apart from each other to gain a better global sense of the LC. Also, we extended previous studies by evaluating the changes in LC curvature for a mean follow-up of more than 3 years. Our results indicate that there is sustained long-term reduction in LC curvature in glaucomatous eyes following successful trabeculectomy. 

Consistent with previous results,^21,26 we found that younger age and larger preoperative LCCI were associated with a greater reduction in LC curvature. This finding is also consistent with previous studies on cadaver eyes from human donors.⁴⁴ The loss of compliance in the posterior scleral region could cause a lack of LC morphologic changes after IOP-lowering treatment in older patients, which is also consistent with longitudinal studies reporting that old age is a risk factor for glaucoma deterioration.^45–47 We propose two possible mechanisms for the association between a larger preoperative LCCI and a larger reduction in the LCCI after trabeculectomy. First, the LC is exposed to two pressures: IOP and intracranial pressure (ICP). The pressure discrepancy between them is referred to as translaminar pressure difference (TLPD, where TLPD = IOP – ICP).⁴⁸ TLPD is considered to play an important role in the pathophysiology of glaucoma.⁴⁹ Larger preoperative LCCI under the influence of TLPD in glaucomatous eyes might suggest less LC stiffness. When TLPD decreases after IOP-lowering treatment, those eyes could demonstrate greater reversal of LC posterior bowing. Also, in light of current limitations with the vertical resolution of imaging technology, a larger preoperative LCCI could contribute to easier detection of LCCI changes. 

In contrast to the current study, previous studies have reported that reductions in LC displacement was also significantly influenced by the magnitude of IOP lowering, which we could not demonstrate with the results of this study.^{19–23,25,29} We considered that the inconsistency could be attributed to the difference in the magnitude of IOP lowering in the studies. In the study by Lee et al.,¹⁹ the mean IOP reduction at 6 months postoperatively was 17.5 mm Hg, but in our study the mean IOP reduction was 10.6 mm Hg. Studies on patients with normal-tension glaucoma that have reported relatively smaller magnitudes of IOP reduction are consistent with our study. Kim et al.²⁶ reported that the reversal of LC curvature was not significantly correlated with the magnitude of IOP reduction in eyes with normal-tension glaucoma after treatment. The smaller magnitude and less variation in IOP reduction could make it difficult to detect a correlation between the magnitude of IOP reduction and LC morphologic changes. Meanwhile, our results showed that lower mean IOP during follow-up examinations was associated with a larger degree of LCCI reduction. This is consistent with previous studies using cadaver eyes that reported increases in LC posterior bowing with increasing IOP.⁵⁰ 

In this study, we also found that the degree of LCCI reduction after trabeculectomy was associated with the baseline refractive error of the eye. In myopic eyes, axial elongation could result in scleral thinning^51,52 and tissue loss,⁵² which may cause structural weakening in myopic eyes and render the LC more susceptible to IOP changes. Lee et al.²⁵ evaluated the changes in LC depth in myopic eyes with POAG 3 months after trabeculectomy. Their results suggest that the degree of reduction in LC depth was not associated with the degree of myopia; however, direct comparison of the results from that study and ours is not feasible. We suspect that variations in methodology may contribute to the conflicting results. The study by Lee et al.²⁵ measured LC depth in different horizontal ONH scans, whereas in our study we measured the LCCI in radial ONH scans. 

Our study has a few limitations. First, due to the retrospective nature of our study, there may be inherent risks of missing data and participation bias. Axial length, which could serve as a better parameter to investigate the association between myopic change and LCCI reduction, was not one of the parameters in our study. We aimed to use the degree of refractive error to reflect participants’ myopic status; however, having data on axial length would be preferable, considering that 13 of the study eyes (32.5%) were pseudophakic. Future prospective observations are needed to address these limitations. Second, the postoperative visits in this study were conducted at a university hospital, resulting in variable follow-up intervals for the study participants. Some patients were only referred for surgical intervention rather than regular glaucomatous follow-ups, which could also have contributed to our relatively limited sample size. Third, our analysis relied on selected B-scans rather than evaluating the curvature of the entire anterior surface of the LC. However, we evaluated six different planes to provide a more comprehensive description of the entire anterior surface. Finally, we conducted the measurements of LCCI manually, which could introduce observer-related bias. To mitigate this bias, we employed a double-masked measurement approach with two independent readers. 

In conclusion, glaucomatous eyes undergoing trabeculectomy demonstrated flattening in LC curvature within 6 months postoperatively, a finding that remained statistically significant over a mean follow-up of 3 years. Greater short-term LCCI reduction was associated with younger age at baseline. Greater long-term LCCI reduction was associated with younger age, lower mean IOP during the follow-up period, greater spherical equivalent refractive error, and preoperative LCCI. 

Acknowledgments

Disclosure: X. Shang; None, N.U. Häner; None, J.-B. Lincke; None, V. Pfeiffer; None, P.A. Gubser; None, M.S. Zinkernagel; Allergan (F), Bayer (F), Boehringer Ingelheim (F), Heidelberg Engineering (F); J.D. Unterlauft; None