Stratus optical coherence tomography study of filtering blebs after primary trabeculectomy with a fornix-based conjunctival flap

Authors


Kazuyuki Hirooka
Department of Ophthalmology
Kagawa University Faculty of Medicine
1750-1 Ikenobe
Miki
Kagawa 761-0793
Japan
Tel: +81 87 891 2211
Fax: +81 87 891 2212
Email: kazuyk@med.kagawa-u.ac.jp

Abstract.

Purpose:  To investigate the relationship between filtering bleb function and Stratus optical coherence tomography (OCT) images.

Methods:  We studied 72 eyes of 65 consecutive patients who had a fornix-based conjunctival flap in primary trabeculectomy with mitomycin C (MMC). Filtering blebs with various types of glaucoma were examined using Stratus OCT. Success rates were defined as intraocular pressure (IOP) ≤ 15 mmHg and IOP reduction ≥ 25% without medication or additional surgery. Success rates among classified groups were compared using Kaplan–Meier survival curves and the log-rank test.

Results:  Blebs were classified into three different categories on the basis of the following Stratus OCT patterns: cystoid type (multiple cysts inside the bleb; 17 eyes), diffuse type (low to high reflective areas that were mixed inside the bleb; 31 eyes) and layer type (medium to high reflective layer inside the bleb; 24 eyes). Success rates in the cystoid-, diffuse- and layer-type blebs were 94%, 97% and 75% (P = 0.02), respectively.

Conclusion:  In trabeculectomy with MMC and a fornix-based conjunctival flap, there is a significant association between the success rates and the postoperative Stratus OCT findings of filtering blebs.

Introduction

Trabeculectomy has become the method of choice in the surgical treatment of patients with glaucoma (Cairns 1968). Filtering surgery success depends mainly on the function of the filtering bleb. For long-term success, the critical factors include not only surgical technique but also age, type of glaucoma (Mills 1981), prior failed filtering surgery and long-term use of topical medication before surgery (Broadway et al. 1994). However, the reason for unsuccessful filtration is often unclear. Frequently, failure is caused by conjunctival scarring and obliteration of the bleb as opposed to internal occlusion of the ostomy. In order to correlate the clinical and functional aspects, many authors have investigated the morphological criteria of the bleb (Vesti 1993; Cantor et al. 2003; Wells et al. 2004). However, in some cases, there is no correlation between the bleb appearance or shape and intraocular pressure (IOP). Clinically, it is not easy to study the reasons for failure related to bleb scarring and the histological changes within the bleb tissue. Unfortunately, this makes it difficult to undertake appropriate therapeutic interventions, because therapy can only be attempted based on identification of the cause of failure.

Intrableb imaging is usually performed by ultrasound biomicroscopy (UBM), a technology that has been developed specifically for the purpose of studying the anterior segment of the eye (Yamamoto et al. 1995; Chiou et al. 1998; Marchini et al. 2001; Roters et al. 2002; Zhou et al. 2006). Optical coherence tomography (OCT) is a recently developed, non-invasive, non-contact technique for imaging of the layered structure of the retina (Huang et al. 1991; Hee et al. 1995). Anterior segment imaging has been reported by several investigators who used either the prototype OCT system (Izatt et al. 1994; Hoerauf et al. 2000; Kalev-Landoy et al. 2007) or the retinal OCT scanner (Muscat et al. 2002). The majority of these studies focused on corneal imaging. OCT has been used to examine filtering blebs after non-penetrating deep sclerectomy (Nozaki et al. 2002). Babighian et al. (2006) reported Stratus OCT images of filtering bleb after trabeculectomy. Savini et al. (2005) reported recently that filtering blebs could be classified into three different categories according to their OCT pattern. However, in the limited number of patients observed, no correlation was noted between their OCT bleb pattern and IOP (Savini et al. 2005). Anterior-segment OCT (AS-OCT) has also been used in internal bleb assessment (Singh et al. 2007). In the current study, we used Stratus OCT (OCT 3; Carl Zeiss Meditec, Dublin, CA, USA) to examine filtering blebs after primary trabeculectomy with a fornix-based conjunctival flap in order to investigate the relationship between bleb function and its Stratus OCT image.

Materials and Methods

Patients

We evaluated 72 filtering blebs of 65 patients who had previously undergone trabeculectomy between October 2003 and October 2005, and who were followed up at the Kagawa University Hospital, Japan. Patients ranged in age from 32 to 86 years [mean ±standard deviation (SD) 67.0 ±11.2 years]. There were 33 male and 32 female patients. All eligible participants received a detailed explanation before both the trabeculectomy and the Stratus OCT. In this study, we did not use slit-lamp-adapted OCT (SL-OCT). All patients signed an informed consent form, in accordance with the Declaration of Helsinki. Twenty-three eyes with primary open-angle glaucoma, 10 with normal-tension glaucoma, four with primary angle-closure glaucoma, 27 with secondary glaucoma caused by uveitis and eight with pseudoexfoliation glaucoma were included in the study.

All patients underwent a standard ophthalmological examination including slit-lamp examination, Goldmann applanation tonometry and binocular fundus examination on day 1, weeks 1 and 2, months 1, 2, 3, 6 and 12 post-operatively, and as necessary.

Trabeculectomy technique

Trabeculectomy was performed by one of the two surgeons (K.H., T.B.). After retrobulbar anaesthesia with lidocaine 2%, the eye was prepared and draped. A corneal traction suture was placed with a 6-0 silk and the fornix-based conjunctival flap was dissected. After the formation of a one-half-thickness scleral flap (approximately 4 × 4 mm), mitomycin C (MMC) was applied with a sponge containing 0.4 mg/ml MMC solution placed under the conjunctival flap for 3–5 min. After the sponge was removed, the area covered by the sponge was irrigated copiously with 250 ml of physiological saline. At the edge of the corneoscleral bed, a block of clear cornea and trabecular meshwork tissue was removed and peripheral iridectomy was performed. The scleral flap was sutured with six or seven 10-0 nylon sutures. The conjunctiva was closed using 10-0 nylon sutures at the edges of the incision, with one or more horizontal mattress sutures placed centrally. The anterior chamber was reformed with balanced salt solution and the wound checked for leaks. A corticosteroid/antibiotic ointment was instilled, followed by a sterile eye patch and shield. Postoperatively, all patients were treated with topical corticosteroid (four times daily) and an antibiotic for 8–12 weeks.

OCT procedure and bleb examination

Stratus OCT examination was performed 12 months after the surgery and before any further glaucoma surgery (including needling). The filtering bleb was visualized on the video monitor and the length of the scanning line was between 5 mm and 7 mm (Fig. 1). All scans were oriented horizontally to the limbus and covered the entire surface of the bleb, from the nasal to the temporal edge. The same examiner (M.T.) took 10 images of a bleb. The images included five sections horizontal to the limbus along with five vertical sections nasal to the temporal region. Focus was adjusted manually on the conjunctiva. Patients were instructed to look downward during the examination. Stratus OCT images for each eye were stored on the instrument’s hard drive. To ensure consistency, a single researcher (K.H.) who was masked toward the clinical features and IOP control analysed all of the Stratus OCT images.

Figure 1.

 The line acquisition protocol was used and the length of the scanning line was between 5 mm and 7 mm. The white line indicates the scanning line.

Bleb appearance was characterized using a slit lamp. The Indiana Bleb Appearance Grading Scale (IBAGS) has been described previously (Cantor et al. 2003). Three parameters are scored: height of the bleb (H), with a scale ranging from H0 (flat) to H4 (high); extent of the bleb (E), with a scale ranging from E (< 1 hr) to E3 (> 4 hr); and vascularity of the bleb, with a scale of V0 (avascular) to V4 (extensive vascularity).

Statistical methods

For the purpose of this study, clinical filtration success was defined in terms of IOP. Surgical success was defined as IOP ≤ 15 mmHg and IOP reduction ≥25% without anti-glaucoma medication. Eyes that required further glaucoma surgery (including needling) or glaucoma medication to lower IOP were considered failures.

Statistical analysis was performed using spss for Windows (SPSS Inc., Chicago, IL, USA). Kaplan–Meier survival analysis and the log-rank test were used to determine success rates. A P-value of < 0.05 was considered statistically significant. Data are presented as means ± SD.

Results

Images were obtained in 72 blebs of 65 patients. Bleb images were obtained easily in all eyes. The Stratus OCT images of the inside of blebs after trabeculectomy with a fornix-based conjunctival flap could be classified into three categories. For each patient, 10 bleb images were obtained; the same category was found to be present in each eye for all of the Stratus OCT images. Cystoid-type blebs (17 eyes) showed multiple cysts inside the bleb (Fig. 2A). Diffuse-type blebs (31 eyes) showed low to high reflective areas that were mixed inside the bleb (Fig. 2B). Layer-type blebs (24 eyes) were characterized by a medium to high reflective layer inside the bleb (Fig. 2C). In the cystoid-, diffuse- and layer-type blebs, the success rates were 94%, 97% and 75% (P = 0.02), respectively (Fig. 3). There were bleb failures for one eye for both cystoid- and diffuse-type blebs, and six bleb failures in the layer-type bleb group. At 1, 2, 3, 6 (two eyes) or 7 months after surgery, a total of six of these eyes required needling. Two eyes had IOPs outside the success criteria range at 6 and 12 months after surgery, and an encapsulated bleb developed in six eyes. In five of these six encapsulated blebs, Stratus OCT indicated the presence of layer-type blebs.

Figure 2.

 Bleb imaging after trabeculectomy with a fornix-based conjunctival flap. (A) Cystoid-type bleb. Optical coherence tomography (OCT) shows the filtering bleb with multiple cysts inside the bleb. (B) Diffuse-type bleb. OCT shows the filtering bleb with low to high reflective areas that were mixed inside the bleb. (C) Layer-type bleb. OCT shows the filtering bleb with a medium to high reflective layer inside the bleb.

Figure 3.

 Results of Kaplan–Meier survival analysis with success defined as intraocular pressure (IOP) ≤ 15 mmHg and IOP reduction ≥ 25% without medication or additional surgery. Success rates in the cystoid-, diffuse- and layer-type blebs were 94%, 97% and 75% (= 0.02) at 12 months, respectively.

There were no significant differences related to glaucoma diagnosis and the Stratus OCT images of the inside of blebs (P = 0.15, χ2 test; Table 1). The relation between bleb characteristic and the Stratus OCT images of the inside of blebs is shown in Table 2. There was no significant difference related to Stratus OCT images and bleb height (P = 0.824, χ2 test), extent (P = 0.754, χ2 test) or vascularity (P = 0.666, χ2 test). Table 3 shows the relation between failure bleb characteristic and Stratus OCT pattern.

Table 1.   Relation between type of glaucoma and Stratus optical coherence tomography (OCT) pattern.
DiagnosisCystoid type (= 17)Diffuse type (= 31)Layer type (= 24)
  1. POAG, primary open-angle glaucoma; NTG, normal-tension glaucoma; PXS, pseudoexfoliation glaucoma; PACG, primary angle-closure glaucoma.

POAG, NTG, PXS81914
PACG310
Secondary glaucoma61110
Table 2.   Relation between bleb characteristics and Stratus optical coherence tomography (OCT) pattern.
OCT patternIndiana bleb appearance grading scale
H0H1H2H3E0E1E2E3V0V1V2V3V4
  1. H, height; E, extent; V, vascularity.

Cystoid type17810287041300
Diffuse type11513212199391720
Layer type0814201167361230
Table 3.   Relation between failure bleb characteristics and Stratus optical coherence tomography (OCT) pattern.
OCT patternIndiana bleb appearance grading scale
H0H1H2H3E0E1E2E3V0V1V2V3V4
  1. H, height; E, extent; V, vascularity.

Cystoid type1000001000100
Diffuse type0100000100010
Layer type0150014110410

Discussion

The long-term success of trabeculectomy is dependent mainly on the development of a functioning filtering bleb. Therefore, the formation and maintenance of functioning blebs are of primary importance. Because clinical bleb evaluation is limited to superficial tissue layers, important data about bleb development may be missed. When using UBM, the structure inside the filtering bleb can be visualized clearly. In addition, features of the UBM images have been shown to have a strong association with filtering bleb function when expressed as the IOP control level (Yamamoto et al. 1995).

AS-OCT was applied for the evaluation of filtering blebs at a single point in time within a widespread period after surgery (2 months to 16years; Singh et al. 2007). In contrast, all blebs except failures were imaged at 12 months after the surgery in our study. In a recent report by Savini et al. (2005), OCT was used to study filtering blebs in a limited number of patients. The authors noted that the blebs could be classified into three different categories according to their OCT patterns. However, their OCT bleb pattern results differed from those noted in the present study. In Savini et al.’s study, data that could have influenced the bleb appearance – such as the site of the conjunctival incision (fornix- or limbus-based) and the Tenon’s capsule excision – were not included. All of our patients received trabeculectomy with a fornix-based conjunctival flap, for which no part of Tenon’s capsule was excised. The desirable bleb morphology after a trabeculectomy is different between limbus-based conjunctival flaps and fornix-based conjunctival flaps. In another study in which paediatric and young adult trabeculectomies were given high doses of MMC, the patients with limbus-based conjunctival flaps were more likely to develop cystic blebs (Wells et al. 2003). This study found that 90% of patients with limbus-based conjunctival flaps developed cystic blebs, whereas this only happened in 29% of patients with fornix-based conjunctival flaps.

The precise pathogenic mechanism involved in scar tissue formation after glaucoma surgery is still not understood completely. However, human Tenon fibroblasts from the subconjunctival space are known to be the central player in wound repair and in the scarring processes after filtrating glaucoma surgery (Khaw et al. 1994). Through proliferation, migration, production and subsequent contraction of extracellular matrix components, human Tenon fibroblasts from scar tissue could be responsible for the filtering failures of the bleb (Khaw et al. 1994; Occleston et al. 1997). The capsule of the encapsulated bleb behaves clinically as a relatively non-contractile tissue. It has been described as being pushed toward the conjunctiva by the aqueous pressure, thereby creating a dome-shaped bleb (Scott & Quigley 1988). This is different from bleb scarring, where flattening of the bleb wall is evident because of myofibroblast contraction (Ariyan et al. 1978). This suggests that non-contractile collagen-producing fibroblasts play the major role in the process of the encapsulated bleb. Similarly, contractile fibroblasts are also the major component in wound healing following filtering surgery. We speculate that the highly reflective walls in blebs might be reflecting the intensity of the intrableb fibroblasts. Because layer-type blebs might have fibroblast proliferation, this type of bleb was seen in eyes with poor IOP control.

MMC is applied during filtering surgery to reduce the risk of bleb failure (Chen 1983). By inhibiting cell proliferation, MMC prevents an excessive healing response and scarring, and thus enhances the success of the procedure. Therefore, we speculate that MMC might reduce the appearance of layer-type blebs.

In conclusion, Stratus OCT is a promising tool in the imaging of morphological changes within the bleb tissue. Without Stratus OCT, it may be difficult to distinguish layer-type blebs from either cystoid- or diffuse-type blebs, because there is no relationship between slit-lamp examinations and Status OCT examinations. Assessment of the morphological changes within the bleb tissue enhances our understanding of the bleb function and may aid clinicians in making decisions regarding postoperative bleb management.

Acknowledgements

This work was supported by Grant-in-Aid for Scientific Research C (20592078).

Ancillary