For the last 40 years, trabeculectomy has been the procedure of choice in glaucoma surgery.1–4 Standard nonpenetrating glaucoma surgery (NPGS) currently consists of different methods, the most popular of which are deep sclerectomy and viscocanalostomy.5,6 The technique of deep sclerectomy was first reported by Kozlov and has since gained acceptance as a viable alternative for the surgical treatment of open-angle glaucoma (OAG). The goal of NPGS was to create a surgical procedure as efficient as trabeculectomy but with less complications.7–9 The main idea behind NPGS is to target the portion of the aqueous outflow pathway responsible for the main resistance to outflow, and to create filtration through trabeculo-Descemet's membrane (TDM).
The uveoscleral pathway was described more than 30 years ago.10 It was also termed unconventional outflow route (as opposed to the conventional or trabecular meshwork outflow) and shown to be responsible for up to 50% of aqueous humor drainage in monkey eyes.10 In uveoscleral outflow, aqueous humor percolates through the extracellular spaces of the ciliary muscle into the suprachoroidal space and exits across the anterior or posterior sclera, through the emissarial canals around the vortex veins, or into the choroidal vessels.11 However, the uveoscleral outflow mechanism is not well understood since currently there is no direct and noninvasive method for its measurement.
Different mechanisms for the reduction of the intraocular pressure (IOP) after NPGS have been proposed including filtration into the subconjunctival space with bleb formation; intrascleral and suprachoroidal flow; unrecognized micropenetration into the anterior chamber; and opening of previously nonfunctional areas of Schlemm's canal.8,12–14
Very deep sclerectomy with collagen implant (VDSCI) is a modification of deep sclerectomy that was devised with the goal of increasing aqueous outflow through the uveoscleral pathway. By dissecting part of the deep sclera, the drainage of aqueous humor into the suprachoroidal space should be enhanced. Aqueous in the suprachoroidal space may hypothetically reach the uveoscleral outflow, and it could also induce a chronic ciliary body detachment thus reducing aqueous production. At the same time, we aimed to reduce dependency on subconjunctival outflow to minimize the size of the subconjunctival filtering bleb and bleb-related discomfort and complications. The primary objective of this study was to compare the IOP lowering effect and safety of the VDSCI with standard DSCI. The secondary objective of this study was to assess the dimensions of the subconjunctival and the suprachoroidal space, and to evaluate the presence of new aqueous drainage pathways under the scleral flap with ultrasound biomicroscopy (UBM).
PATIENTS AND METHODS
The glaucoma unit, Jules Gonin Eye Hospital, University of Lausanne, Switzerland.
The trial design followed the CONSORT guidelines.15 After the approval of the Ethical Committee of the University of Lausanne, the patients were enrolled consecutively. The purpose and nature of the trial was explained in detail to all participants and their informed consent was obtained. From July 2003 to August 2004, 50 eyes of 48 patients with medically uncontrolled primary or secondary OAG underwent nonpenetrating glaucoma surgery. Patient demographic data are shown in Table 1. Randomization was performed using Random Figure Tables designating the surgery to either DSCI (25 eyes) or VDSCI (25 eyes). Two eyes of 2 patients were randomly assigned to each of the 2 study arms, which explains the fact that despite 48 patients included in the trial, each group consisted of 25 patients. The patients were masked to the procedure they received.
Inclusion criteria were patients with uncontrolled primary or secondary OAG, despite maximum tolerated medication, and no known allergy to collagen. Exclusion criteria were age less than 18 years, fertile women without adequate contraceptive device and breastfeeding mothers. Medically uncontrolled glaucoma was defined as well-documented progression of visual field defects under maximal medical therapy and increasing glaucomatous optic nerve head cupping documented photographically. Before surgery, all patients underwent best-corrected visual acuity (BCVA) assessment (Snellen chart at 5 m), slit lamp biomicroscopy, gonioscopy, Goldmann applanation tonometery (Haag-Streit, Bern, Switzerland), pachymetry, visual field testing using the G 1 program of the Octopus 101 (Interzeag AG, Schlieren, Switzerland), and mydriatic fundus examination.
All surgeries were performed by 1 experienced surgeon using retrobulbar anesthesia. A superior rectus muscle suture was placed. The conjunctiva and Tenon's capsule were then opened in the upper fornix. A one-third scleral thickness limbus-based scleral flap, measuring 5×5 mm was performed. A second deep scleral flap measuring 4×4 mm was dissected. When the Descemet's membrane had been exposed for 1 mm, the second scleral flap was excised.
At this stage of surgery, the very deep scleral dissection was performed in the VDSCI group. In the posterior quadrants of the deep scleral bed, 2 very deep flaps (each 1.5×1.5 mm) of the remaining 5% to 10% of sclera were excised and the choroid was exposed. A thin bridge of deep scleral tissue was left between the 2 flaps to prevent a possible choroidal prolapse. (Fig. 1) At that last stage of the procedure, which was equal in the 2 groups, the juxtacanalicular trabeculum and Schlemm's endothelium were removed and diffuse percolation of aqueous through the remaining TDM could be observed. A collagen implant was placed in the center of the scleral bed and secured to the sclera with a single 10/0 nylon suture. The superficial scleral flap was then repositioned over the implant and secured. The procedure is described in more detail elsewhere.16
Mitomycin C (MMC) was applied during surgery in 9 DSCI patients compared with 8 VDSCI patients (P=NS). Criteria for intraoperative application of MMC were patients at high risk for postoperative fibrosis (ie, age below 60 years, patients of African origin, previous history of conjunctival surgery, long-standing history of glaucoma treatment, previous uveitis, or trauma). MMC was administered after the dissection of the superficial scleral flap by applying a MMC 0.02% soaked sponge for the duration of 30 seconds and washed out with balanced salt solution.
Postoperatively, the same examinations (except for visual field assessment) were performed at day 1, at week 1, at months 1, 2, 3, 6, 9, 12, 18, and 24 months of follow-up. Main outcome measures were the success rate based on IOP, visual acuity, number of postoperative medications, and early and late complications. Visual field assessment was performed at a minimum of 6 months after surgery. In addition, UBM was performed at 3 and 12 months after surgery.
Complications were defined as follows: hyphema was considered present when erythrocytes were seen in the anterior chamber. Hypotony was defined as a postoperative IOP of less than 5 mm Hg for more than 4 weeks. Anterior chamber depth was clinically assessed in comparison with the other eye and was considered shallow when there was irido-corneal touch in the periphery and flat when there was lens-corneal touch as seen with biomicroscopy. Anterior chamber inflammation was considered present when flare could be seen. In postoperative follow-up, cataract was either observed as a direct consequence of filtering surgery, considered surgery-related cataract, or appeared progressively and was therefore considered physiologic cataract progression. When the filtering bleb at any postoperative visit was encysted or fibrotic, needling with or without a subconjunctival injection of MMC was administered.
Goniopuncture with the Nd: YAG laser (Lasag AG, Thun, Switzerland) was performed when percolation of aqueous humor at the TDM was considered to be insufficient either due to a shallow filtration bleb or elevated IOP. Laser goniopuncture techniques and results have previously been reported.17
The Humphrey UBM 840 system (Humphrey Instruments Inc, San Leandro, CA) was used to provided high frequency (50 MHz) ultrasonic scan images. Scans were performed by an experienced investigator, using the technique developed by Pavlin et al.18 Radial and transverse sections of the sclerectomy area were explored. Biometric measurements of the surgical site were performed with electronic calipers installed in the instrument. The following variables were assessed: (a) the presence and maximum length, height and width of the subconjunctival filtering bleb and of the intrascleral lake and (b) the possibility of suprachoroidal drainage as an alternative aqueous pathway. The presence of suprachoroidal or supraciliary low-echogenic area in the sclera extending posterior to the ciliary body was considered evidence of possible suprachoroidal drainage.
Criteria for Success
Complete success was defined as an IOP of ≤18 mm Hg and a percentage drop of at least 20%, achieved without medication; qualified success was applied for the same IOP limitation but with or without glaucoma medication.
The sample size (50 eyes) was chosen to assure a power of 90% or more in detecting at least a 2.5 mm Hg difference between groups, and a standard deviation of 2.5 mm Hg, with a 2-sided α error of 5%. Student t tests were used to compare continuous variables between the 2 groups such as IOP differences. To compare the 2 groups on day 1 and after 1 year, an analysis of variance with preoperative data as a covariate was performed. Preoperative data were included in the model, to adjust for a possible influence of the preoperative IOP on the later trend of the IOP. Measurements of IOP were performed by evaluators other than the surgeon. Kaplan-Meier survival analysis (log-rank test) comparing the success rates of the 2 groups was done. All P values reported are 2-tailed and significance was defined as P<0.05. Statistical analysis was done using commercially available statistical software program Stata 9.0 (Stata Corporation, College Station, TX).
The mean follow-up period was 18.6±5.9 months (median: 17.0) for the VDSCI group, and 18.9±3.6 months (median: 16.4) for the DSCI group (P=NS). One patient died during the follow-up. Three eyes (all in VDSCI group) experienced rupture of TDM during surgery and were converted to a deep sclerectomy with a perforation, in which the deep scleral block was removed and a peripheral iridectomy was performed. In the DSCI group 1 eye required a second surgical intervention (trabeculectomy) due to uncontrolled postoperative IOP control. This case was considered a surgical failure. All of these cases were included in the analysis, following the rule that once randomized, all patients should be analyzed.
Mean preoperative IOP was 22.4±7.4 mm Hg (range: 15 to 42) for VDSCI and 20.4±4.4 mm Hg (range: 14 to 30) for DSCI eyes (P=NS). Those patients with low IOP values were included due to documented progression of disease despite maximal tolerated medication. The mean postoperative IOP was 3.9±2.2 mm Hg (VDSCI) and 6.3±4.3 mm Hg (DSCI) at day 1 (P<0.05), 9.4±5.2 mm Hg (VDSCI) versus 12.5±4.0 mm Hg (DSCI) at month 1 (P=0.05), and 12.2±3.9 mm Hg (VDSCI) versus mm Hg 13.3±3.4 (DSCI) at month 24 (P=NS) (Fig. 2, Table 2). No significant differences of IOP were found preoperatively and after 1 year. At the last visit, the complete success rate (defined as an IOP of ≤18 mm Hg, achieved without medication) was 57% in VDSCI and 62% in DSCI eyes (P=NS). Qualified success (IOP ≤18 mm Hg with or without medication) was achieved in 68% VDSCI and 72% DSCI eyes (P=NS) (Fig. 3).
BCVA dropped on the first postoperative day from a mean of 0.86±0.2 preoperatively to a mean of 0.63±0.3 in the VDSCI group, whereas in the DSCI group, BCVA dropped from a mean of 0.89±0.3 preoperatively to a mean of 0.73±0.3 on the first postoperative week (P=NS). Visual acuity returned to preoperative levels 2 weeks after surgery and remained stable over the entire follow-up period; the mean was 0.81±0.2 (VDSCI) and 0.88±0.3 (DSCI) (P=NS) at 12 months.
The mean number of medications per patient was reduced from 2.16±0.8 to 0.24±0.5 (P<0.001) in the VDSCI group, and from 2.12±1.0 to 0.20±0.4 (P<0.001) in the DSCI group. The between-group difference of reduction in medications was statistically not significant. Goniopuncture with the Nd:YAG laser was performed on 10 patients (40%) in the VDSCI group compared with 14 patients (54%) in the DSCI group (P=NS). Mean time of goniopuncture was 8.0±6.8 (VDSCI) and 7.5±4.9 (DSCI) months (P=NS) (Table 3). There were no significant postoperative complications in this series. The 2 interventions were similar in terms of postoperative complications (Tables 3, 4). No shallow or flat anterior chamber, no bleb-related endophthalmitis, and no surgery-induced cataract were observed in either group. There was 1 case of malignant glaucoma in a DSCI patient, which underwent conservative management and resolved without further intervention.
The need for postoperative MMC injections was significantly higher in the DSCI group (10 cases vs 5 in VDSCI, P<0.05). The dose of subconjunctival MMC injection was 0.02%. The appearance of blebs remained unchanged in most patients in both groups and improved after MMC injection to a more diffuse and less tense one in 3 VDSCI and 1 DSCI eye. When Cox regression analysis with intraoperative and postoperative MMC injections as confounding factors was performed, no significant difference could be shown between the 2 groups. There were 23 diffuse blebs in the VDSCI versus 22 in the DSCI, 2 cystic (VDSCI) versus 3 (DSCI) and no avascular blebs in either group.
Figure 4 shows the typical UBM appearance of the intrascleral bleb and new hypoechoic supraciliochoroidal spaces after VDSCI. UBM examinations at 3 months postoperatively showed a mean volume of the subconjunctival filtering bleb of 4.4±4.5 mm3 (VDSCI) and 8.1±9.5 mm3 (DSCI) (P=0.32). The mean volume of the intrascleral space was 5.7±3.7 mm3 (VDSCI) and 5.4±2.2 mm3 (DSCI) (P=0.82). At 12 months, the corresponding values were 3.9±4.2 (VDSCI) and 6.8±7.5 (DSCI) (P=0.426) for the subconjunctival filtering bleb and 5.2±3.6 mm3 (VDSCI) and 5.4±2.9 mm3 (DSCI) (P=0.90) for the intrascleral space. The presence of a supraciliochoroidal low-echoic area in the sclera was observed in 32% (VDSCI) and 28% (DSCI) eyes at 3 months (P=0.35), and in 36% (VDSCI) and 24% (DSCI) eyes at 12 months (P=0.48).
The results of this prospective, randomized trial indicate that the new method of very deep sclerectomy is a safe variation of standard DSCI. In our study, we did not observe any significant differences in terms of success rate and complications between the 2 groups. Contrary to our expectations, we could not show a significantly higher level of outflow facility and increased presence of an uveoscleral pathway mechanism with VDSCI. However, the fact that the need for postoperative MMC injections was significantly lower in the VDSCI group could indicate a potential advantage over standard DSCI. The number of cases in this series is small with relatively short follow-up and hence conclusions should be drawn with these limitations in mind.
The role of NPGS next to trabeculectomy in the management of medically uncontrolled glaucoma has been the subject of much debate in recent years. Although different studies show a similar pressure-reducing effect of NPGS with a significantly better safety profile, to date, no adequately powered randomized controlled trial has compared both methods head-to-head.19,20 A major benefit of NPGS lies in a significantly reduced size of the filtering bleb, documented over the long-term follow-up.8,9,14 As a consequence, the rate of postoperative blebitis or endopthalmitis is extremely rare and to our knowledge, no such report has yet been published, whereas after trabeculectomy incidence of bleb-related endophthalmitis has been reported to be around 2%.21
The mechanism of aqueous drainage in NPGS remains controversial. Different reports have suggested that after the removal of juxtacanalicular tissue and inner wall of Schlemm's canal, where the majority of flow resistance is located, drainage may either be subconjunctival, intrascleral, eventually through Schlemm's canal, or through enhanced uveoscleral outflow in the suprachoroidal space.13 Our rationale for studying the described variation to standard deep sclerectomy was to create a mechanism of iatrogenic “bypass cyclodialysis” which would provide a communication between the anterior chamber and the uveoscleral pathway. A hypoechoic area indicative of effusion in the suprachoroidal space may be seen in normal eyes. In our experience, however, this anatomic variant is much smaller and less frequent than has been observed in eyes after deep sclerectomy. It is postulated that the suprachoroidal pathway, contrary to the trabecular meshwork outflow pathway, is pressure-independent.11 Therefore, being able to enhance this route might further improve the surgical success once IOP is lowered in a first step through the classic pathway.
In this study, the difference between the 2 groups in terms of reduction of IOP at a mean of 18.7 months and the number of patients requiring medication postoperatively was not significant. Modern glaucoma surgery is often performed in eyes with preoperative IOPs of around 20 mm Hg. To reflect this fact, we defined complete success rate as an IOP ≤18 mm Hg and a percentage drop of at least 20%, without medication. The complete and qualified success rates of the VDSCI were statistically similar to DSCI. When a different cutoff level of IOP ≤16 mm Hg, which corresponds to the desired target pressures for most patients, was used, there was again no significant difference between the 2 groups (Fig. 3).
The complication rate of both procedures was in accordance with published literature.8 There were 3 cases of intraoperative perforation in the VDSCI eyes that were immediately transformed into a deep sclerectomy with perforation. These complications should not be attributed to the variation in technique as they involved the anterior part of the surgical site that is identical in both procedures.
In this study, UBM was used to evaluate the morphology of the postoperative filtering area. UBM findings at 3 and 12 months showed a reduction in the volume of subconjunctival bleb and also a stable intrascleral space in both groups. Suprachoroidal effusion was present in 36% VDSCI and 24% DSCI eyes at 12 months. We were expecting to find more suprachoroidal hypoechogenic areas and smaller filtering blebs in the VDSCI group, but that was not revealed in the study. Although the results tended to point to this direction, the difference was not statistically significant.
Different UBM studies have analyzed the morphologic changes after deep sclerectomy. Chiou and coworkers12 were the first to show the presence of suprachoroidal hypoechogenic areas after deep sclerectomy by UBM, which was present in 44% of cases in their series of 9 patients 1 month after surgery. In a follow-up study with 45 eyes, the same group found signs of suprachoroidal effusion in 51% at 18 months.13 Marchini et al22 studied the outcome of deep sclerectomy with reticulated hyaluronic acid implants in 30 patients and found an association between the presence of the intrascleral space, the supraciliary hypoechogenic area, and IOP lowering.22 In their series, a suprachoroidal hypoechoic area was present in 60% of eyes after a mean of 11.4 months. Another study assessing 43 eyes for 1 year observed hypoechogenic areas in the suprachoroidal space in 44% of the eyes.14 Khairy et al23 in a similar study did not detect a hypoechogenic suprachoroidal area in any of the 22 eyes after 12 months of follow-up.23 No scleral implant was used in this study. With the exception of the latter, all of the above-mentioned studies could show a significant correlation between the presence of a hypoechoic suprachoroidal space and lower IOP. This finding was confirmed in this study.
Contrasting with these findings were 2 studies that analyzed eyes after viscocanalostomy. These studies could not support an association between an enhanced uveal pathway and surgical success. This could, however, be due to the different outflow mechanism of this technique that relies more on reestablishing the physiological anatomic pathway through the Schlemm's canal. One recent study looked at UBM findings after trabeculectomy.24 The authors used a variation of the procedure in which the scleral spur was excised to remove the barrier between the anterior chamber and the suprachoroidal space. This prospective series of 28 eyes showed the presence of suprachoroidal fluid in 12.5% of cases after 6 months and an association with lower postoperative IOP.
At a mean follow-up of 18.7 months, there was no statistically significant difference between very deep sclerectomy with a collagen implant and standard deep sclerectomy with a collagen implant in terms of IOP, complete and qualified success rates, and reduction in number of medications. UBM examinations showed no statistically significant increase in amount of suprachoroidal effusion between the 2 groups. The need for postoperative MMC injections was significantly lower in the VDSCI group. In light of this potential benefit and the similar safety profile, we conclude that VDSCI could be an alternative to standard penetrating surgery by potentially decreasing the complications and discomfort related to the subconjunctival bleb. However, at this stage of research, more patients with a longer follow-up are needed to comprehensively assess the safety and efficacy of this new procedure.
1. Krasnov MM. Externalization of Schlemm's canal (sinusotomy) in glaucoma. Br J Ophthalmol. 1968;52:157–161.
2. Zimmermann TJ, Kooner KS, Ford VJ, et al. Effectiveness of nonpenetrating trabeculectomy in aphakic patients with glaucoma. Ophthal Surg. 1984;15:44–49.
3. Zimmermann TJ, Kooner KS, Ford VJ, et al. Trabeculectomy versus Nonpenetrating trabeculectomy: a retrospective study of two procedures in phakic patients with glaucoma. Ophthal Surg. 1984;15:734–740.
4. Kozlov VI. Nonpenetrating deep sclerectomy with collagen. Ophthal Surg. 1990;3:44–46.
5. Stegmann RC. Visco-canalostomy: a new surgical technique for open angle glaucoma. An Inst Barraquer Spain. 1995;25:229–232.
6. Kozlov VI, Bagrov SN, Anisimova SY, et al. Non-penetrating deep sclerectomy with collagen. Eye Microsurg (Russian). 1995;3:44–46.
7. Watson PG, Jakeman C, Ozturk M, et al. The complications of trabeculectomy: a 20-year follow-up. Eye. 1990;4:425–438.
8. Shaarawy T, Mansouri K, Schnyder C, et al. Long-term results of deep sclerectomy with collagen implant. J Cataract Refract Surg. 2004;30:1225–1231.
9. Mermoud A, Schnyder CC, Sickenberg M, et al. Comparison of deep sclerectomy with collagen implant and trabeculectomy in open-angle glaucoma. J Cataract Refract Surg. 1999;25:323–331.
10. Bill A. The aqueous humor drainage mechanism in the cynomolgus monkey (Macaca irus) with evidence for unconventional routes. Invest Ophthalmol. 1965;4:911–919.
11. Alm A, Weinreb RN, eds. Uveoscleral Outflow. London: Mosby-Wolfe; 1998.
12. Chiou AGY, Mermoud A, Hédiguer SEA, et al. Ultrasound biomicroscopy of eyes undergoing deep sclerectomy with collagen implant using an ultrasound biomicroscope. Br J Ophthalmol. 1996;80:541–544.
13. Chiou AG, Mermoud A, Underdahl JP, et al. An ultrasound biomicroscopic study of eyes after deep sclerectomy with collagen implant. Ophthalmology. 1998;105:746–750.
14. Kazakova D, Roters S, Schnyder CC, et al. Ultrasound biomicroscopy images: long-term results after deep sclerectomy with collagen implant. Graefes Arch Clin Exp Ophthalmol. 2002;240:918–923.
15. Moher D, Schulz KF, Altman DG, CONSORT. The CONSORT statement: revised recommendations for improving the quality of reports of parallel group randomized trials. BMC Med Res Methodol. 2001;1:2.
16. Shaarawy T, Karlen M, Schnyder C, et al. Five-year results of deep sclerectomy with collagen implant. J Cataract Refract Surg. 2001;27:1770–1778.
17. Mermoud A, Faggioni R, Schnyder CC, et al. Nd-Yag goniopuncture after deep sclerectomy with collagen implant. Invest Ophthalmol Vis Sci. 1996;37:1167.
18. Pavlin CJ, Sherar MD, Foster FS. Subsurface ultrasound microscopic imaging of the intact eye. Ophthalmology. 1990;97:244–250.
19. El Sayyad F, Helal M, El-Kholify H, et al. Nonpenetrating deep sclerectomy versus trabeculectomy in bilateral primary open-angle glaucoma. Ophthalmology. 2000;107:1671–1674.
20. Ambresin A, Shaarawy T, Mermoud A. Deep sclerectomy with collagen implant in one eye compared with trabeculectomy in the other eye of the same patient. J Glaucoma. 2002;11:214–220.
21. Freedman J, Gupta M, Bunke A. Endophthalmitis after trabeculectomy. Arch Ophthalmol. 1978;96:1017–1018.
22. Marchini G, Marraffa M, Brunelli C, et al. Ultrasound biomicroscopy and intraocular-pressure-lowering mechanisms of deep sclerectomy with reticulated hyaluronic acid implant. J Cataract Refract Surg. 2001;27:507–517.
23. Khairy HA, Atta HR, Green FD, et al. Ultrasound biomicroscopy in deep sclerectomy. Eye. 2005;19:555–560.
24. Martínez-Belló C, Capeáns C, Sánchez-Salorio M. Ultrasound biomicroscopy in the diagnosis of supraciliochoroidal fluid after trabeculectomy. Am J Ophthalmol. 1999;128:372–375.