Skip Navigation LinksHome > August 2014 - Volume 23 - Issue 6 > Descemet Membrane Detachment After Canaloplasty: Incidence...
Journal of Glaucoma:
doi: 10.1097/IJG.0b013e318279ca7f
Original Studies

Descemet Membrane Detachment After Canaloplasty: Incidence and Management

Jaramillo, Arley MD; Foreman, Julie MD; Ayyala, Ramesh S. MD, FRCS, FRCOphth

Free Access
Article Outline
Collapse Box

Author Information

Glaucoma Service, Department of Ophthalmology, Tulane University Medical Center, New Orleans, LA

Supported in part by the Tulane Glaucoma Research Fund.

Presented in part as a poster at the Association for Research in Vision and Ophthalmology Annual Meeting, Fort Lauderdale, FL, May 2009.

Disclosure: R.S.A. was a consultant to iScience Interventional. The remaining authors declare no conflict of interest.

Reprints: Ramesh S. Ayyala, MD, FRCS, FRCOphth, Glaucoma Service, Department of Ophthalmology, Tulane University Medical Center, 1430 Tulane Avenue, SL-69, New Orleans, LA 70112 (e-mail:

Received March 12, 2012

Accepted October 9, 2012

Collapse Box


Purpose: To report the incidence and management of Descemet membrane detachment (DMD) after canaloplasty.

Methods: Review of all patients who developed DMD after canaloplasty at Tulane Glaucoma Services was performed. Visual acuity, intraocular pressure, and number of glaucoma medications up to 1 year of follow-up were included in the analysis.

Results: The incidence of DMD was 7.4% (12 eyes of 162). Eighty-three percent (10/12) of the DMDs involved the inferior quadrants and measured <3 mm. Intracorneal hemorrhage within DMD occurred in 58% (7/12), whereas 42% (5/12) developed DMD with intracorneal viscoelastic (Healon GV) alone. Two patients had large detachments measuring 5 to 6 mm extending into the visual axis. DMD resolved completely with or without drainage except for 1 patient who developed corneal decompensation, needing penetrating keratoplasty.

Conclusions: DMD with or without intracorneal hemorrhage is not an infrequent complication of canaloplasty and can occasionally lead to corneal decompensation.

Canaloplasty is a new surgical glaucoma procedure that involves viscodilation of the Schlemm canal with the placement of an intracanalicular tension suture. Compared with traditional surgeries, this has the advantage of independence from bleb and bleb-related problems.1–4 Rare complications of this surgery that have been previously reported include transient hypotension, hyphema, and Descemet membrane detachment (DMD).1–4 As canaloplasty is a relatively new procedure, the natural history, incidence, and management of DMDs after canaloplasty is not well documented. We report the natural history, incidence, and management of patients with DMD after canaloplasty in our series.

Back to Top | Article Outline


Institutional Review Board approval was obtained from Tulane University for this study. A retrospective chart review was performed on all patients who had canaloplasty at Tulane University Health Sciences Center from January 2007 to July 2011. All surgeries were performed by a single surgeon (R.S.A). A total of 162 eyes (of 115 patients) were reviewed. All patients who had DMD were identified. The following data were recorded from their charts: Preoperative and postoperative intraocular pressure (IOP), number of glaucoma medications, visual acuity (VA), and incidence of DMD. Postoperative data were gathered for postoperative day 1, 1 week, 1, 3, 6, 9, and 12 months.

Back to Top | Article Outline
Operative Technique

The canaloplasty procedure was performed in the standard manner previously described.1 Briefly, after conjunctival dissection at the 12 o’ clock limbus, a 5×5 mm partial thickness (50%) scleral flap was dissected followed by a 4×4 mm inner scleral flap at 95% depth. The inner flap dissection was carried forward until Schlemm canal was unroofed. The dissection was carried into clear cornea to create a 0.3-mm Descemet window followed by the removal of the inner scleral flap. This was followed by viscodilation of the canal using sodium hyaluronate 1.4% (Healon GV; Advanced Medical Optics Inc., Santa Ana, CA) with the help of an ophthalmic microcatheter (iTrack-250A; iScience Interventional Inc., Menlo Park, CA). The quantity of Healon GV that was injected was controlled by quarter turn of the special syringe provided by the company. Healon GV was injected every 2 to 3 clock hours to a total of 8 times in all cases. Inward traction was applied to the dilated Schlemm canal by inserting a 10-0 prolene suture (Ethicon Inc., Somerville, NJ) with the help of the microcatheter, which was then tightly tied in a loop with a slip knot. The scleral flap was secured back to the sclera with 5 10-0 nylon sutures to create a water-tight closure. The conjunctiva was then secured to the limbus with 10-0 vicryl sutures (Ethicon Inc.).

DMD was noticed at the time of the surgery in all cases. If DMD involved significant blood mixed with viscoelastic (meaning, mostly hemorrhagic than viscoelastic), it was drained within 1 week. Drainage was performed at the slit lamp by making a stab incision over the detachment. The details are as follows. Informed consent was obtained. A drop of topical lidociane 1% was applied to the eye followed by Betadine prep. The lids were kept apart using a lid speculum. The patient was positioned at the slit lamp. A 1- to 2-mm corneal incision was made using a 15-degree blade at the site of maximum detachment. The incision was carefully continued until the detachment pocket was entered. The incision was kept open using a jeweler’s forceps letting the blood mixed with viscoelastic out. The incision was closed with a single interrupted 10-0 nylon suture. All patients were maintained on prednisolone acetate 1% every 2 hours, ketorolac 0.5% qid and gatifloxacin 0.3% qid for 1 week, and prednisolone acetate 1% qid for 1 month.

Back to Top | Article Outline
Statistical Analysis

The IOP data were summarized by mean values and SD, and VA was expressed in terms of the mean and SE in logarithm of the minimum angle of resolution units. Sex, race, eye (left, right) and previous surgery were compared between groups using the Fisher exact test; the Student t test was used to assess age differences. Baseline IOP, VA, and number of medications were compared between the 2 groups. Repeated-measures analysis of variance with Bonferroni adjustment was used to compare the 2 groups with respect to IOP and VA from preoperative baseline to 12 months. Statistical analysis was performed using the SPSS software package (version 18.0; SPSS Inc./IBM, Chicago, IL). Two-tailed P<0.05 was considered significant.

Back to Top | Article Outline


A total of 162 eyes (of 115 patients) that underwent canaloplasty were reviewed. Twelve patients developed DMD after canaloplasty (7.4%) (12 eyes of 162) (DMD group). All 12 patients with DMD were followed for a minimum of 12 months. Only132 eyes of the 150 eyes in the non-DMD group were followed for a minimum of 12 months, and thus, the postoperative data (vision, medications and IOP) were limited to these 132 eyes. The demographics and baseline characteristics between those that developed DMD (DMD group) and those that did not develop this complication (non-DMD) is listed in Table 1. No differences were found with respect to age (P=0.98), sex (P=0.99), race (P=0.94), right versus left eye (P=0.77), diagnosis (P=1.00), preoperative VA (P=0.99), and number of glaucoma medications (P=0.90) between the 2 study groups (Table 1).

Table 1
Table 1
Image Tools

Number of glaucoma medications and the VA at baseline and at 12 months between the 2 groups were comparable between the 2 study groups (Table 1). The preoperative and postoperative IOP between the 2 groups was compared at various time points up to 12 months. No differences were found between the 2 groups. The average preoperative IOP for patients who experienced DMD was 20.5 mm Hg and that in the non-DMD group was 21.2 mm Hg (P=0.75). The mean IOP between the 2 groups at 12 months was 12.3 mm Hg (DMD) and 14.5 mm Hg (non-DMD) (P=0.17) with no significant differences between the 2 groups at any time points (Table 1).

The average age of the DMD group was 68. 6 years, and there were 7 male and 5 female patients. Three patients were phakic and 9 patients were pseudophakic at the time of DMD. The DMD occurred most commonly in the inferior quadrants [inferonasally (5) or inferotemporally (5)]. In 2 patients, it occurred in the superiotemporal region. All patients developed the detachment during viscodilation of the Schelmm canal with Healon GV. Five eyes developed DMD with intracorneal viscoelastic (Healon GV) without hemorrhage and measured between 1 and 2 mm in size. Seven of the 12 eyes developed an intracorneal hemorrhage with DMD (Figs. 1A, B) and measured between 2 and 3 mm in size. The detachment measured 1 to 2 mm in size in 5 eyes and 2 to 3 mm in size in 5 eyes. Two patients had large detachments measuring 5 to 6 mm extending into the visual axis.

Figure 1
Figure 1
Image Tools

The DMD with viscoelastic alone resolved without any intervention by 1 month after operation, in all patients, without any squelae. Of the eyes that developed intracorneal hemorrhage, the DMD and hemorrhage resolved without intervention by 6 weeks in 3 eyes. The blood in the detachment area faded over a 3-month period as a gray-white deposit before disappearing completely. In the rest of the 3 eyes, the detachment resolved with intervention at the slit lamp by 10 weeks. The decision to intervene was based on the size (>3 mm) and the proximity to or the involvement of the visual axis.

The slit-lamp surgical intervention was performed between 1 day and 1 week after the surgery. The last patient in this group was an 86-year-old African American man who developed DMD toward the end of surgery. The DMD extended 3 mm from the inferonasal quadrant and was mixed with blood. It did not involve the visual axis at the end of surgery, and so, no attempt was made to drain it. On postoperative day 1, the DMD was still the same size. However, at 1 week, the patient returned with a 6-mm detachment with blood extending into the visual axis (Fig. 2). This was drained at the slit lamp on that day.

Figure 2
Figure 2
Image Tools

The detachment still persisted over the next week involving the visual axis. The detachment was treated with external drainage through a stab incision and 100% intracameral injection of air bubble in the operating room. Even though the detachment resolved, the patient developed corneal decompensation over the next 3 months and underwent penetrating keratoplasty (PKP) at 6 months. At the last follow-up (3 y after the PKP), the patient had excellent IOP control (11 mm Hg on timolol) and vision returned to preoperation levels (20/200).

Back to Top | Article Outline


DMD with or without intracorneal hemorrhage is a known complication after cataract surgery.5 DMD after canaloplasty is a relatively new entity and the incidence has been reported to range from 1% to 6.25%.1–4 In our study, the incidence was 7.4%. DMD could be related to excessive amounts of Healon GV injection into the Schlemm canal during the viscodilation portion of the surgery. This explanation may have some truth to it as the primary surgeon (R.S.A.) noticed excessive injection of the Healon GV by inexperienced nursing staff in at least 2 cases. We experienced only 1 case of DMD during the first year followed by 2 to 4 cases/year during the subsequent years using the same technique. It has previously been postulated that there is an increased risk of DMD in canals that have adhesions between the inner and outer wall. Most of the detachments in our series occurred in the inferior quadrants. This could be because of one of 2 reasons. As the Healon GV is injected from the entry point at or near 12 o’ clock limbus, the Healon GV may have reached a critical mass in the inferior quadrants leading to the detachment of the Descemet membrane from the Schwalbe line. The intraluminal resistance during the procedure may surpass the strength of the connection between Descemet membrane and Schwalbe line resulting in a detachment of Descemet membrane The other possibility is that there could be an anatomic predisposition or congenital weakness in the inferior quadrants. Palmiero et al6 reported a case with bilateral inferonasal detachment after canaloplasty and postulated that this may be because of anatomic predisposition. Both these cases resolved without intervention over 3 months. The last possibility could be related to the surgical technique used in the current study. The technique that was used in the study was the same that was previously published. The most likely variable could be the amount of viscoelastic that was injected. The exact amount of the viscoelastic that should be injected into the canal in total and the amount that comes out of the company supplied syringe per quarter turn are unknown and deserve future research.

All but one of our cases resolved without any permanent sequelae. In 58% of our cases (7/12), blood was mixed with Healon GV in the detachment. In 3 of these cases, we chose to drain at the slit lamp because of the presence of significant quantity of blood mixed with Healon GV and the size of the detachment (>3 mm).7 The DMD resolved completely in 2 of the 3 eyes. One case progressed to corneal decompensation despite air bubble injection into the anterior chamber and drainage of blood. This was an 86-year-old man who developed progressive corneal edema, ultimately needing a PKP.

Limitations of the present study include all the inherent weaknesses of a retrospective study, including possible selection bias and limited patient population. This study analyzed the surgical results of a single surgeon (R.S.A); thus, the results may not be applicable to everyone. Some factors that may have relevance to the present study but were not studied include endothelial cell count and the number of times the viscoelastic was injected in every case.

In conclusion, DMD with or without intracorneal hemorrhage can occur in up to 7% of the patients after canaloplasty. Most cases occur in the inferior quadrants and resolve without any permanent sequelae. Occasionally, DMD can result in central corneal decompensation needing PKP and should be recognized as a potential serious adverse event after canaloplasty.

Back to Top | Article Outline


1. Ayyala RS, Chaudhry AL, Okogbaa CB, et al..Comparison of surgical outcomes between canaloplasty and trabeculectomy at 12 months’ follow-up.Ophthalmology.2011;118:2427–2433.

2. Grieshaber MC, Fraenkl S, Schoetzau A, et al..Circumferential viscocanalostomy and suture canal distension (canaloplasty) for whites with open-angle glaucoma.J Glaucoma.2011;20:298–302.

3. Lewis RA, von Wolff K, Tetz M, et al..Canaloplasty: circumferential viscodilation and tensioning of Schlemm’s canal using a flexible microcatheter for the treatment of open-angle glaucoma in adults: interim clinical study analysis.J Cataract Refract Surg.2007;33:1217–1226.

4. Shingleton B, Tetz M, Korber N.Circumferential viscodilation and tensioning of Schlemm canal (canaloplasty) with temporal clear corneal phacoemulsification cataract surgery for open-angle glaucoma and visually significant cataract: one-year results.J Cataract Refract Surg.2008;34:433–440.

5. Mulhern M, Barry P, Condon P.A case of Descemet’s membrane detachment during phacoemulsification surgery.Br J Ophthalmol.1996;80:185–186.

6. Palmiero PM, Aktas Z, Lee O, et al..Bilateral Descemet membrane detachment after canaloplasty.J Cataract Refract Surg.2010;36:508–511.

7. Gismondi M, Brusini P.Intracorneal hematoma after canaloplasty in glaucoma.Cornea.2011;30:718–719.


canaloplasty; Descemet membrane detachment; penetrating keratoplasty

© 2014 by Lippincott Williams & Wilkins.


Article Level Metrics

Search for Similar Articles
You may search for similar articles that contain these same keywords or you may modify the keyword list to augment your search.