Evaluating subconjunctival bleb function after trabeculectomy using slit-lamp optical coherence tomography and ultrasound biomicroscopy

Trabeculectomy has been the standard surgical treatment for glaucoma since its introduction in the 1960s.¹ The outcome depends mostly on the resulting intraocular pressure. The description of the morphology and function of blebs are usually based on the clinician's subjective judgment of conjunctival vascularization, bleb height and width, the presence or absence of subconjunctival microcysts and cystic diffusion.^2-4Representative bleb grading systems in current use include the Kronfeld grading system,⁵ Indiana Bleb Appearance Grading Scale (IBAGS),⁶ and Moorfields Bleb Grading System (MBGS).² However, these systems do not include intrableb structure imaging criteria. In this study, we imaged intrableb structures using slit-lamp-adapted optical coherence tomography (SL-OCT; Heidelberg Engineering GmbH, Heidelberg, Germany) and ultrasound biomicro- scopy (UBM; Tomey, D1000, Japan) and morphologically classified the blebs, to compare the efficiency of the two techniques in assessing bleb function.

Study subjects
Ethical approval for this study and agreement by all patients were obtained from the Shanghai Jiao Tong University Affiliated Sixth People's Hospital. Each subject signed an agreement of participation in this study that was approved by Sixth People's Hospital Affiliated to Shanghai Jiao Tong University. The protocol of the study adhered to the tenets of the Declaration of Helsinki and was approved by the Local Ethics Committee. We randomly chose 29 eyes of 21 male patients and 40 eyes of 32 female patients who underwent glaucoma filtering surgery (including 41 trabeculectomy surgeries and 28 phacotrabeculectomy surgeries) from more than 300 glaucoma filtering surgery in our hospital between 2002 and 2006. The patients ranged in age from 29 to 83 years (mean 67.7±8.3 years). The eyes included 37 with acute angle closure glaucoma, 22 with chronic angle closure glaucoma, and 10 with primary open angle glaucoma. The interval between surgery and the SL-OCT/UBM examination ranged from 6 months to 4 years (mean 22.2±13.7 months).

Standard of functioning filtering blebs
Based on the method of Singh et al⁷ successful blebs intraocular pressure were defined as follows. For blebs with a preoperative intraocular pressure (IOP) >18 mmHg, with or without ocular hypotensive medication, or for blebs with a preoperative IOP ≤18 mmHg with ocular hypotensive medication, success was defined as a last recorded IOP ≤18 mmHg without topical glaucoma medication. For blebs with a preoperative IOP ≤18 mmHg without ocular hypotensive medication, a 20% drop in IOP with no ocular hypotensive medication was accepted as success. According to these standards, this study included 69 eyes of 53 subjects, including 29 eyes in 21 male patients and 40 eyes in 32 female patients.

SL-OCT test
For SL-OCT, the patient was asked to look down, and the operator elevated the upper lid gently to expose the limbus, iris incision, bleb and surrounding bulbar conjunctiva as fully as possible, taking care to avoid putting pressure on the globe or bleb. Scanning was performed vertically to the limbus, with a 15-mm linear scan at the center of the bleb. Bleb morphology was evaluated, including bleb height, bleb wall thickness, intra-bleb reflectivity, subconjunctival fluid space, suprascleral fluid space, scleral flap thickness, the route under the scleral flap and patency of the internal ostium. All images were measured using built-in software. We noted the presence of a low reflective region under the conjunctiva, thickened conjunctiva, highly reflective region around the scleral flap, and patent internal ostium, as well as the filtering route.

UBM test
For UBM, the patient was under topical anesthesia in the supine position using an eye cup filled with saline as a coupling medium. Scanning was performed perpen- dicularly to the limbus in the region of the trabeculectomy. The following factors were identified: filtering bleb height, internal cavity, internal ostium after trabeculectomy and route beneath the scleral flap. We classified the filtering blebs using the classification of Yamamoto et al.⁸ In addition, the SL-OCT and UBM images were compared.

Statistical analysis
Statistical analysis was performed using SPSS for Windows (version 13.0). A Fisher′s exact test was used to compare the sensitivity for predicting if a functioning bleb differed significantly between SL-OCT and UBM. A Fisher′s exact test was also used for morphological analysis of the trabeculectomy blebs based on SL-OCT. P <0.05 was considered statistically signiﬁcant.

According to the standard of Singh et al,⁷ 45 (65.2%) of the 69 blebs were functioning and 24 (34.8%) were non-functioning. In all subjects, the preoperative IOP ranged from 18.9 to 81.0 mmHg (mean IOP, (48.6±8.5) mmHg). The preoperative IOP with ocular hypotensive medication was between 10.1 and 37.1 mmHg (mean IOP (16.8±4.6) mmHg). One year after surgery, the IOP ranged from 7.6 to 29.6 mmHg (mean IOP (17.9±5.3) mmHg).

The morphous and function of the filtering blebs tested by SL-OCT
In this study, we observed intra-bleb structures using SL-OCT. Most of the functioning blebs had a thickened bleb wall or low reflective lacunae of different shapes beneath the conjunctiva. The non-functioning blebs usually exhibited adhesions, which showed high reflectivity, between the conjunctiva and scleral flap, and had a non-thickened bleb wall and an occluded internal ostium. Based on the SL-OCT images, we classified the 69 filtering blebs into four categories: diffuse filtering, cystic, encapsulated and flattened (Figures 1–5).

view in a new window

Figure 1. Diffuse filtering bleb: (A) image obtained by SL-OCT. SCFS: Subconjunctival fluid space; FS: fluid space above scleral flap; SF: scleral flap; R: route under scleral flap; c: conjunctive; Co: cornea; Ir: Iridectomy; S: Sclerotomy; (B) image obtained by UBM; (C) front segment image.
Figure 2. Cystic bleb: (A) image obtained by SL-OCT. SCFS: Subconjunctival fluid space; FS: fluid space above scleral flap; c: conjunctive; Co: cornea; Ir: Iridectomy; S: Sclerotomy; I: Iris (B) image obtained by UBM; (C) front segment image.
Figure 3. Encapsulated bleb: (A) image obtained by SL-OCT. c: conjunctive; FS: fluid space above scleral flap; SF: scleral flap; Co: cornea; Ir: Iridectomy; S: Sclerotomy; (B) image obtained by UBM; (C) front segment image.
Figure 4. Flattened bleb: (A) image obtained by SL-OCT. Co: cornea; Ir: Iridectomy; S: Sclerotomy; I: Iris. (B) image obtained by UBM; (C) front segment image.
Figure 5. Diffuse filtering bleb: (A) image obtained by SL-OCT. SCFS: Subconjunctival fluid space (“stripe sign”); SF: scleral flap; R: route under scleral flap; Co: cornea; Ir: Iridectomy; S: Sclerotomy; (B) image obtained by UBM; (C) front segment image.

Diffuse blebs
Diffuse blebs were characterized by the presence of several subconjunctival microcysts or lacunae that showed low or moderate reflectivity. They had diffusely thickened conjunctiva, and the scleral flap, suprascleral fluid space, a route under the scleral flap and the internal ostium were clearly identified (Figure 1A). Of the 69 blebs, 39 (56.5%) were diffuse blebs exhibiting these features. Using the standard of Singh et al,⁷ 35 blebs were successful and four blebs were non-functioning.

Cystic blebs
A cystic bleb (Figure 2A) was also a functioning bleb. The bleb wall was relatively thin, usually less than 0.2 mm. The cystic blebs were of varying height and were composed of a large lacuna or several small or medium lacunae above the scleral flap, some of which were fused with the lacunae beneath the conjunctiva. The route beneath the scleral flap was not clearly displayed. There were three (4.3%) cystic blebs, all of which were successful according to Singh et al⁷ and had postoperative IOPs between 12.2 and 16.8 mmHg after 1 year.

Encapsulated blebs
An encapsulated bleb (Figure 3A) was a failed bleb in which the fluid-filled suprascleral space was surrounded by highly reflective tissue. In contrast to a successful cystic bleb, an encapsulated bleb usually had no subcon- junctival lacuna, and the scleral and conjunctival tissues were not clearly distinguished, possibly even adherent. In general, there was no diffuse thickened bleb conjunctiva, but rather the entire bleb wall was thick. The bleb wall in the two present cases exceeded 0.4 mm. The suprascleral fluid space was surrounded by highly reflective dense connective tissue, and the scleral and conjunctival tissues were not adherent. The route beneath the scleral flap and the internal ostium were usually patent. The two encapsulated blebs (2.9% of total blebs) were non-functioning based on Singh et al.⁷ The postoperative IOPs were 25.7 to 29.6 mmHg after 1 year.

Flattened blebs
A flattened bleb (Figure 4A) was similar to an encapsulated bleb and contained no lacuna beneath the conjunctiva or above the scleral flap. The entire bleb was low, with a thick wall. The adhesive scleral and conjunctival tissues were highly reflective, similar to the scleral reflective signal. The route beneath the scleral flap was usually absent, although the internal ostium could be patent. Microcysts beneath the conjunctiva, lacunae above the scleral flap and the route beneath the scleral flap could not be identified using SL-OCT. There were 25 (36.2%) flattened blebs: 7 functioning and 18 non-functioning, according to Singh et al.⁷

Morphological analysis of the trabeculectomy blebs based on SL-OCT
Several significant differences were noted between the functioning and non-functioning bleb groups. We observed a thick bleb wall, with or without microcysts, beneath the conjunctiva in 35/45 (77.8%) eyes (mean IOP, (14.3±3.6) mmHg) in the functioning bleb group but in only 4/24 (16.7%) eyes (mean IOP, (22.9±3.2) mmHg) in the non-functioning bleb group. The difference was significant (P=0.000, Fisher′s exact test). A low reflective region around the scleral flap was present in 26/45 (57.8%) eyes (mean IOP, (12.4±3.7) mmHg) in the functioning bleb group but in only 5/24 (20.8%) eyes (mean IOP, (22.3±3.1) mmHg) in the non-functioning bleb group (P=0.005，Fisher′s exact test). A highly reflective region above the scleral flap was found in 12/45 (26.7%) eyes (mean IOP, (15.8±2.1) mmHg) in the functioning bleb group and in 14/24 (58.3%) eyes (mean IOP, (23.4±2.8) mmHg) in the non-functioning bleb group (P=0.018, Fisher′s exact test). A route beneath the scleral flap was absent in 20/45 (44.4%) eyes (mean IOP, (15.1±2.7) mmHg) in the functioning bleb group and in 19/24 (79.2%) eyes (mean IOP, (23.6±3.2) mmHg) in the non-functioning bleb group (P=0.010，Fisher′s exact test). The five eyes with an occluded internal ostium (mean IOP, (24.3±2.6) mmHg) all belonged to the non-functioning bleb group (P=0.004, Fisher′s exact test).

The morphous and function of the filtering blebs tested by UBM
The same blebs were also classified based on the UBM images ( Figure 1B, Figure 2B, Figure 3B, Figure 4B, Figure 5B), using the method of Yamamoto et al.⁸ They were classified into 17 type L blebs, 19 type H blebs, 4 type E blebs and 29 type F blebs. Using the standard of Singh et al, of 17 type L blebs, 15 blebs were successful and 2 blebs were non-functioning. Of 19 type H blebs, 15 blebs were successful and 4 blebs were non-functioning. Of 4 Type E blebs, 2 blebs were successful and 2 blebs were non-functioning. Of 29 Type F blebs, 13 blebs were successful and 16 blebs were non-functioning.

Comparison of the results between the SL-OCT test and the UBM test
We determined the sensitivity and specificity for each imaging method. A comparison of the images obtained with the two techniques showed that SL-OCT had a higher resolution than UBM. In some failed Type F blebs classified using UBM, SL-OCT showed a microcyst beneath the conjunctiva or a thick bleb wall. We observed that a “stripe sign” was shown in the Figure 5A.

With SL-OCT, only three successful blebs were classified as flattened blebs, and the sensitivity for predicting a successful bleb was 92.7% (38/41 eyes), with reference to the standard of Singh et al.⁷ By contrast, four non-functioning blebs were classified as diffuse blebs, giving a specificity of 83.3% (20/24 eyes). With UBM, 15 successful blebs were classified as Type F or Type E filtering blebs, and the sensitivity for predicting a functioning bleb was 66.7% (30/45 eyes), with reference to the standard of Singh et al.⁷ Six non-functioning blebs were classified as Type L or Type H blebs, giving a specificity of 75.0% (18/24 eyes). The sensitivity for predicting a functioning bleb differed significantly between the two groups (P=0.003, Fisher′s exact test).

Since UBM was introduced in the 1990s, it has been used to evaluate filtering blebs after trabeculectomy, and it was proven to predict bleb function with good specificity and sensitivity.^9-13 Compared with UBM, SL-OCT has a higher axial resolution. SL-OCT can qualitatively reveal bleb structures such as the scleral flap, internal ostium, or cystic lacuna and can semi-quantitatively determine bleb height, bleb size, bleb wall thickness and scleral flap thickness.^14,15

In this study, we used SL-OCT to observe internal bleb structures and classified the blebs into four categories. These findings were compared with the bleb classification based on UBM and the classification system of Yamamoto et al.⁸ However, the classification system of Yamamoto was less sensitive for predicting a functioning bleb. This might be attributable to the high resolution of SL-OCT, which allows the recognition of structures such as microcysts beneath the conjunctiva, a low reflective region adjacent to the scleral flap, and a thick bleb wall in some functioning Type H blebs.

We classified the blebs with these structures as diffuse blebs, which greatly increased the sensitivity for predicting a functioning bleb. In this type of bleb, although the aqueous humor absorbs the reflection signal at a wavelength of 1310 nm, the sclera reflection signal below the filtering zone is weakened significantly; Theelen et al¹⁶ defined this as the “shading” phenomenon. In addition to the shading phenomenon, Theelen et al¹⁶ also occasionally observed that a “striping” phenomenon within Tenon's layer in functioning blebs was associated with good functionality of the filtration site, suggesting the presence of abundant drainage channels in successful filters. The absence of this structure in non-functioning encapsulated or flattened blebs confirmed the hypothesis. Further study is needed to prove whether the length or height of the shading or striping region and the width of the internal ostium are linearly related to IOP control.

In addition, four diffuse blebs in our study were classified as non-functioning blebs according to the method of Singh et al, perhaps because of insufficient filtration. The aqueous humor filtration flow rate did not reach a level that reduced the IOP below 18 mmHg, and thus medication was still necessary. Slit-lamp microscopy detected obvious conjunctival vascularization, vascular engorgement and tense conjunctiva in these four blebs.

Using UBM, Yamamoto et al⁸ classified the blebs with a cystic fluid-filled space surrounded by a thin bleb wall with high signal reflectivity as Type E blebs, and the IOP was uncontrolled in 50% of these blebs. It had no specificity for predicting a successful bleb, perhaps because the identification of Type E blebs could not distinguish between functioning cystic blebs and non-functioning encapsulated blebs. We identified the blebs with a thin bleb wall (usually less than 0.2 mm) and a large cavity or several small or medium lacunae above the scleral flap as cystic blebs; these often had an indistinct route beneath the scleral flap. Encapsulated blebs were characterized by a large suprascleral fluid space surrounded by dense connective tissue with a high reflectivity signal, which indicated that the aqueous drainage had failed because the aqueous outflow was impeded by the surrounding dense fibrous tissue. The bleb wall in both cases exceeded 0.4 mm, which concurs with the data of Leung et al (0.46–0.52 mm).¹⁷ There was usually no lacuna beneath the conjunctiva, with adhesion between the conjunctiva and scleral tissue, and an occluded or patent internal ostium was present. Savini et al¹⁸ reported 6 cases that have the similar characterize. Savini found the cases are characterized by multiple, elevated and communicating cavities under a thin layer of conjunctiva. Irregular septa of connective tissue with low-to-medium reﬂectivity could be seen under the bleb wall, which were named type B blebs.

The high resolution of SL-OCT made it possible to observe lacunae and microcysts beneath the conjunctiva or the low reflective region above the scleral flap. Aqueous humor drainage in lacunae beneath the conjunctiva resulted in good IOP control. In Savini's research,¹⁸ 7 cases were reported with the similar characteristics. Savini found the cases are characterized by the presence of several small, irregular and ﬂat ﬂuid-ﬁlled spaces inside a thick wall, which were named type C blebs. In addition, three flattened blebs were classified as functioning blebs according to the method of Singh et al, perhaps because a functioning trabecular meshwork became patent again after surgery, allowing control of the IOP.

This study detected the presence of a suprascleral fluid space or the route beneath the scleral flap, and the patency of the internal ostium was correlated with bleb function. In this study, all five blebs with occlusion of the ostium were non-functioning filtering blebs. From the SL-OCT images, it was possible to obtain evidence of non-functioning blebs, which might aid the clinician in determining postoperative bleb management.

In conclusion, SL-OCT is a useful tool for observing anterior segment structures, especially the morphological features inside blebs that cannot be seen with a slit-lamp microscope, after glaucoma filtration surgery. Compared with UBM, SL-OCT has increased sensitivity for predicting a successful bleb because it has greater resolution. The non-contact approach of SL-OCT makes it possible to avoid iatrogenic injury, which allows the evaluation of blebs at any time following surgery. SL-OCT offers prompt and objective measures for assessing the outcome of glaucoma filtration surgery or the need for a second operation.

1. Cairns JE. Trabeculectomy. Preliminary report of a new method. Am J Ophthalmol 1968; 66: 673-679.

2. Wells AP, Crowston JG, Marks J, Kirwan JF, Smith G, Clarke JC, et al. A pilot study of a system for grading of drainage blebs after glaucoma surgery. J Glaucoma 2004; 13: 454-460.

3. Richter CU, Shingleton BJ, Bellows AR, Hutchinson BT, O'Connor T, Brill I. The development of encapsulated filtering blebs. Ophthalmology 1988; 95: 1163-1168.

4. Vesti E. Filtering blebs: follow up of trabeculectomy. Ophthalmic Surg 1993; 24: 249-255.

5. Kronfeld FC. The chemical demonstration of transconjunctival passage of aqueous after antiglaucomatous operations. Am J Ophthalmol 1952; 35: 38-45.

6. Cantor LB, Mantravadi A, WuDunn D, Swamynathan K, Cortes A. Morphologic classification of filtering blebs after glaucoma filtration surgery: the Indiana Bleb Appearance Grading Scale. J Glaucoma 2003; 12: 266-271.

7. Singh M, Chew PT, Friedman DS, Nolan WP, See JL, Smith SD, et al. Imaging of trabeculectomy blebs using anterior segment optical coherence tomography. Ophthalmology 2007; 114: 47-53.

8. Yamamoto T, Salcuma T, Kitazawa Y. An ultrasound biomicroscopic study of filtering blebs after mitomycin C trabeculectomy. Ophthalmology 1995; 102: 1770-1776.

9. Pavlin CJ, Harasiewicz K, Sherar MD, Foster FS. Clinical use of ultrasound biomicroscopy. Ophthalmology 1991; 98: 287-295.

10. McWhae JA, Crichton ACS. The use of ultrasound biomicroscopy following trabeculectomy. Can J Ophthalmol 1996; 31: 187-191.

11. Avitabile T, Russo V, Uva MG, Marino A, Castiglione F, Reibaldi A. Ultrasound biomicroscopic evaluation of filtering blebs after laser suture lysis trabeculectomy. Ophthalmologica 1998; 21: S17-S21.

12. Chiou AG, Mermoud A, Underdahl JP, Schnyder CC. An ultrasound biomicroscopic study of eyes after deep sclerectomy with collagen implant. Ophthalmology 1998; 105: 746-750.

13. Marchini G, Marraffa M, Brunelli C, Morbio R, Bonomi L. Ultrasound biomicroscopy and intraocular-pressure-lowering mechanisms of deep sclerectomy with reticulated hyaluronic acid implant. J Cataract Refract Surg 2001; 27: 507-517.

14. Müller M, Hoerauf H, Geerling G, Pape S, Winter C, Hüttmann G, et al. Evaluation of filtering blebs using slit-lamp adapted 1310 nm optic coherence tomography (OCT). Ophthalmologe 2002; 99 Suppl: S72.

15. Müller M, Pape S, Hoerauf H, et al. OCT evaluation of functioning filtering blebs—1 year results. Ophthalmologe 2004; 101 Suppl: S277.

16. Theelen T, Wesseling P, Keunen JEE, Klevering BJ. A pilot study on slit lamp-adapted optical coherence tomography imaging of trabeculectomy filtering blebs. Graefe's Arch Clin Exp Ophthalmol 2007; 245: 877-882.

17. Leung CK, Yick DW, Kwong YY, Li FC, Leung DY, Mohamed S, et al. Analysis of bleb morphology after trabeculectomy with Visante anterior segment optical coherence tomography. Br J Ophthalmol 2007; 91: 340-344.