Descemet's membrane macroperforation during interface irrigation in big bubble deep anterior lamellar keratoplasty.

This report describes the macroperforation of descemet's membrane (DM) during irrigation of the interface following successful removal of stromal tissue in big bubble (BB) deep anterior lamellar keratoplasty (DALK). A 42-year-old woman with keratoconus underwent a BB DALK procedure. After successful formation of Type 2 bubble and removal of stromal tissue, interface irrigation was performed to remove residual viscoelastic. This led to a macroperforation of DM, and the case was converted to penetrating keratoplasty. To the best of our knowledge, this is the first report of such a complication in the literature. Recognition of a 2 bubble formation during pneumodissection should alert the surgeon to a high risk of DM rupture. We advise against the removal of DM from the donor in such cases as an added measure of safety.

Introduction

Deep anterior lamellar keratoplasty (DALK) is a technique used to treat a variety of corneal stromal pathologies such as keratoconus, scars, dystrophies, and degenerations and is thought to be the gold standard approach for anterior corneal pathology in eyes with a healthy endothelium.[1]

Case Report

A 42-year-old woman was seen in clinic with advanced right keratoconus and best-corrected visual acuities (VA) of 20/400 in the right eye and 20/20 in the left eye with a manifest refraction of +5.50 −8.00 × 65 in the right eye and +3.50 −1.25 × 80 in the left eye. Orbscan topography (Bausch and Lomb Inc., Rochester, NY, USA), could not capture any images of the right eye due to the advanced nature of the disease but the apical thickness was found to be 438 μm in the left eye. There was no history of acute hydrops nor did the patient have any other past ocular or medical history of note.

Because of the severity of keratoconus, the patient was unable to wear a contact lens. After discussing treatment options, risks, benefits, and alternatives, the patient elected to undergo corneal transplantation.

Surgical procedure

The procedure was performed under general anesthesia. An 8.00 mm Hessburg Barron (JedMed Instrument Co., St. Louis, MO) suction trephine was used for partial thickness trephination of the host cornea up to an approximate depth of 60%. A 27-G disposable needle bent at 90° was advanced in the paracentral stroma, and 1.0cc of air was injected into the corneal stromal tissue resulting in the formation of Type 2 big bubble (BB) [Figure 1a]. A small bubble of air was injected into the anterior chamber to confirm the formation of the BB [Figure 1b]. After a small opening was made, Vannas’ scissor was used to divide the stromal tissue into four quadrants, and each quadrant was excised baring descemet's membrane (DM) completely [Figure 1c]. After complete excision of stromal tissue, interface irrigation was carried out to remove all remaining viscoelastic material from the interface [Figure 1d]. During irrigation a macroperforation of DM occurred and the procedure was converted to penetrating keratoplasty.

Figure 1

Consecutive images of a big bubble deep anterior lamellar keratoplasty. During the procedure, Type 2 bubble was created using a 27-G bent needle (note the peripheral formation of the bubble with a clear margin). (a) The small bubble technique confirms the presence of the big bubble. (b) Central stromal tissue was removed successfully with no perforation noted. (c) Irrigation of the surgical interface was performed to remove residual viscoelastic, and this led to a macroperforation. (d) Final image after conversion to penetrating keratoplasty and suturing of a full thinness donor lenticule with 16 interrupted 10-0 nylon sutures. (e)

An 8.25 mm full-thickness donor lenticule was then secured over the recipient bed with 16, 10-0 monofilament nylon interrupted sutures [Figure 1e]. Postoperatively, the patient was given 2-hourly prednisolone acetate 1% (Predforte, Allergan, New Zealand) and 4-hourly chloramphenicol 0.5% eye drops (Chlorsig, Aspen Pharma Pty Ltd, Australia). The postoperative period was uneventful, and at the 9 months’ postoperative review the graft was clear, and the patient achieved a best-corrected VA of 20/25 with a manifest refraction of +3.00 −3.00 × 55 [Figure 2].

Figure 2

Postoperative image of the right eye 9 months following penetrating keratoplasty

Discussion

The BB DALK technique provides a planned, safe, and consistent removal of stromal tissue by injection of air deep into the stroma. The technique has a steep learning curve, and DM perforation can occur during any step of surgery, including air injection into deep stroma, dissection of host stroma, and even suturing of the donor button.

While the BB DALK technique was initially believed to lead to a “bare” DM, recent studies have suggested that pneumodissection can create three distinct BB morphologies.[2]

Type 1, where the air separates the predescemet's layer from the deep stroma creating a large central bubble (8–9 mm)

Type 2, wherein the DM is separated from the posterior surface of the predescemet's layer

Mixed, where both bubble types can coexist either partially or completely.

Type 2 BB only occurs in 6%–19%[3] of cases following pneumodissection and is seen more commonly in older patients, thin corneas, and stromal scarring.[4] The distinction between the different types of BBs is important as the reported rate of DM perforation in eyes with Type 2 bubble is much higher than that of eyes with a Type 1 bubble. In one study, there was a more than 10-fold increase in the rate of DM perforation in eyes with a Type 2 bubble where the DM is not supported by the predescemat layer (7% vs. 86%).[4] This is unsurprising when one considers the much lower bursting pressure of Type 2 bubble when compared to Type 1 bubble (0.6 bar and 1.45 bar, respectively).[2]

While no clear surgical strategy has been defined in the literature to deal with the occurrence of Type 2 bubble; it is imperative that all surgical manoeuvres are slow and calculated to reduce the risk of DM perforation. Our case highlights the fragility of the DM penumodissected after a Type 2 bubble formation and its susceptibility to rupture with the mildest of forces. We advocate that DM should not be removed from the donor lenticule in the event of Type 2 bubble formation. This is because a macroperforation can occur unexpectedly at any stage of surgery even during irrigation of the interface, a step that is strongly advocated if viscoelastic has been used to prevent potential pseudo double anterior chamber formation. While the removal of the DM from the donor in BB, DALK has been postulated to reduce the incidence of endothelial rejection, reduce the incidence of interface haze, improve postoperative contrast sensitivity and encourage healing at the donor–recipient interface, it has not been shown to offer improved VA, refractive outcomes nor endothelial cell counts.[5] More importantly, by not removing the donor endothelium, no tissue is lost in the event of DM perforation or in the presence of a double anterior chamber.

In summary, this case highlights the importance of distinguishing between different types of BBs intraoperatively and exercising extreme caution in the case of Type 2 bubble formation.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.