The Big Bubble Technique

Alaa M. Eldanasoury, MD; Sherif Said Tolees, FRCS; and Harkaran S. Bains, BHSc Oph

Anterior lamellar keratoplasty (ALK) involves a partial- or full-thickness resection of the corneal stroma that is replaced by donor tissue. Numerous techniques have been developed for ALK including manual free hand dissection,¹ microkeratome-assisted ALK,^2,3 and deep anterior lamellar keratoplasty (DALK). DALK is a full-thickness, or near–full-thickness, stromal resection down to, or close to, Descemet’s membrane (DM). Common indications for DALK include keratoconus and other corneal ectasias, stromal dystrophies, stromal scars and opacities, corneal ulcers, and corneal perforations. Corneal pathologies affecting the endothelium are contraindicated for DALK.

Many techniques^1–10 have been described for DALK, with the ultimate goal in all being to preserve the recipient’s healthy endothelium and replace only pathologic stromal tissue. The advantage of this approach includes a significant mitigation of the risk of endothelial rejection and late endothelial failure in cases with stromal pathologies and healthy endothelium. In 1985, Archila ⁴ described the injection of air above the DM to facilitate the excision of all corneal layers except DM. Sugita and Kondo⁵ described fluid dissection to maximize the removal of diseased corneal stroma. Melles⁶ described a method to visualize the corneal thickness by filling the anterior chamber with air so that the depth of an instrument relative to the posterior corneal surface was visible during surgery. Jacob^11,12 described her modified technique of pre-Descemetic DALK (PD-DALK) for primary treatment of acute hydrops to avoid scarring and obtain visual, topographic, structural, and biomechanical improvement and decreased dependence on contact lens, while at the same time avoiding comorbidity of 2 surgeries and the risk of scarring over the visual axis necessitating a penetrating keratoplasty (PK).

Many modifications in techniques have been published, including viscoelastic dissection,7 combined air and fluid dissection,⁸ and deep parenchyma detachment from the corneal limbus.⁹ Anwar and Teichmann¹⁰ described a technique for full dissection of the stroma from DM by injecting a big air bubble in the cleavage plane between DM and the deep stroma inducing a central Descemet’s detachment, facilitating complete stromal excision, and reducing the risk of perforation. This technique is called big bubble DALK. A recent report¹³ of the big bubble technique demonstrated that cleavage takes place between the posterior stroma and a newly described pre-Descemet’s layer composed of 5 to 8 collagen lamellae.

BIG BUBBLE DALK

ALK techniques can be divided into 2 main categories based on the depth of dissection: a PD-DALK that leaves variable layers of stroma adherent to DM at the end of the dissection, and the Descemetic procedure (Descemetic DALK [D-DALK]), involving complete stromal excision baring DM or the pre-Descemet’s layer (Figure 5-1).

The big bubble technique is the most popular surgical method to obtain a D-DALK,¹⁴ which allows the cleavage separation between DM and posterior stroma by forceful injection of air into the deep stroma. The overall success of this technique in separating the DM from the posterior stroma is approximately 70% to 80% worldwide.¹⁵

If the big bubble is not obtained, a manual intrastromal dissection has to be performed. The stroma is excised layer-by-layer until the plane of the DM or the pre-Descemetic layer is reached (D-DALK) or a regular deep stromal plane, close to the endothelium is created (PD-DALK)^15,16 (Figure 5-2).

This chapter provides a comprehensive step-by-step description of the big bubble technique for DALK as originally described by Anwar and Teichmann¹⁰ and performed by the authors (over 5000 cases). This chapter guides corneal surgeons who are beginning this procedure or early in their surgical learning curve. The technique described here will highlight the many advantages of DALK with the big bubble technique and surgical pearls are presented for mitigating common difficulties encountered during the learning curve.

The patient’s eye is prepared and draped in a sterile manner. The center for trephination is marked on the recipient cornea, midway between the geometric center of the cornea and the center of the entrance pupil. The diameter for trephination is chosen to include all or most of the pathologic corneal tissue leaving at least 1.0 mm of clear peripheral recipient cornea for suturing. With a suction trephine, a partial-thickness trephination of the recipient cornea is performed to leave between 50 and 150 μm of corneal thickness. After releasing the suction, the wound is inspected for depth. A depth of approximately 300 μm makes the following steps easier. We prefer the use of a limbal suction trephine with a central obturator because it creates a more vertical trephination wound compared to corneal suction trephines (Figures 5-3 and 5-4 and Video 5-1).

A 27-gauge needle, bent at a 60° angle 4 mm from its tip is introduced bevel down through the bottom of the wound in the central cornea. Many air injection cannulas are commercially available that forego the need to bend needles. The needle has to be in the posterior third of the stroma, which is easy to achieve with an adequate partial trephination. With the needle in place, air is slowly injected into the deep stroma. If the needle is positioned at the proper depth, with more stromal tissue anteriorly than posteriorly, the air will dissect its way between stromal lamellae, creating an area of emphysema before following the path of least resistance, to the anatomic plane between the stroma and pre-Descemet’s membrane. This leads to a bullous Descemet detachment, the air bubble in the proper plane can be seen as a well-defined circular bubble raising the central cornea. The big air bubble usually forms between the pre-Descemetic layer and the deep stroma (Type 1 bubble). Less commonly, air dissects its way between the DM and the pre-Descemetic layer (Type 2 bubble). Rarely, 2 bubbles are formed, one in each plane (Type 3 bubble), this is a combination of Type 1 and 2. This is described later in the chapter.

If a big air bubble fails to form and there is relatively high resistance as the air is injected, air will disseminate into the corneal stroma creating surgical emphysema in the superficial corneal tissue. If this occurs, air injection is stopped, the needle withdrawn and reintroduced at a deeper level in an adjacent region of clear cornea. Air injection can be repeated until a large bubble is formed or the cornea becomes totally opaque preventing visualization; in this case, an ALK is performed removing opaque stromal tissue and air is injected at a deeper level as described later in the chapter.

Some surgeons described the use of a sharp-tip dissector with a round shaft to facilitate the lamellar dissection by creating a lamellar track for the 27-gauge blunt air injection cannula. According to proponents of this technique, it yields a statistically higher percentage of successful big bubble creation and makes the maneuver more manageable compared to air injection with a needle.17 In our experience, we did not find a significant difference in the rate of successful bubble creation with the sharp needle vs the use of a dissector and cannula.

Once a big air bubble is obtained, a side port (with a slight vertical orientation to prevent piercing the big bubble) is created in the very peripheral cornea to drain aqueous and lower the pressure in the anterior chamber. Anterior keratectomy is then performed, resecting the anterior portion of the stroma (corneal debulking), this can be easily performed with a round tip blade. At this stage, the large air bubble becomes more evident as the central cornea above the bubble is firm over a soft globe. This is a hallmark that the air bubble is in the correct plane. At this stage the surgeon should also identify the type of bubble formed. Here we describe a typical DALK procedure with the more common Type 1 bubble. Dealing with Type 2 and Type 3 bubbles will be described under the “Surgical Challenges” section later in this chapter.

The donor tissue is cut from the endothelial side using the same diameter blade as used for the recipient. Peeling Descemet and endothelium from the donor tissue can be performed using a dry surgical sponge and/or non-toothed forceps (Figure 5-5). In some cases, the DM with the endothelium can be easily peeled in one layer; this usually occurs with corneal tissue from older donors. In corneal tissue from younger donors (45 years old or younger) DM-endothelium complex can be removed in pieces with a dry sponge. Care should be taken to prevent stromal injury during removal of the endothelium. Watson et al¹⁸ described the use of trypan blue for easier identification of DM and endothelium.

Suturing Donor Graft

The donor tissue is placed over the recipient bed. A 12 or 16 radial marker coated with gentian violet is used to mark both the donor and the recipient tissue, 4 cardinal 10-0 sutures are used to fix the donor in place and 12 or 16 anti-torque running bites are used to secure the donor tissue in place. Suture tension is adjusted based on intraoperative keratoscopy to minimize postoperative astigmatism. Interrupted sutures can be also used according to the surgeon’s preference; however, in the authors’ experience running sutures are more advantageous for adjusting the tension on the graft intraoperatively as well as during the early postoperative period (Video 5-2).

Postoperative Medications

This includes topical antibiotics and steroids. We prescribe topical antibiotics; Vigamox (moxifloxacillin) until complete closure of the epithelium and topical steroids (Pred forte; prednisolone acetate 1%) 4 times/day for 1 month and then gradually tapered over 6 weeks. We leave the continuous sutures in place for 12 months postoperatively to ensure wound healing and stability. In cases where the sutures become loose or attract blood vessels, they are replaced.

During the learning curve, a number of intraoperative challenges can occur. The main challenge is to avoid perforation of DM. The incidence of DM perforation during DALK ranges from 4% to 39.2%5,8,19–21 and appears to depend on surgeon experience, indications for keratoplasty, and surgical techniques. Keratoconus patients are more prone to DM ruptures than patients with other corneal diseases.²² Microperforation rates seem to be highest with manual layer-by-layer dissection (26.3%) and lowest with the Anwar big bubble technique (5.48%). Perforation of DM can occur during any step of surgery, including trephination, air injection, stromal excision, and donor suturing.

Perforation During Trephination

If the trephination is too deep and anterior chamber penetration occurs in more than 2 clock hours, completion of the procedure is extremely difficult and conversion to PK is recommended, especially during the learning curve. If a small penetration occurs, the site or penetration can be sutured and big bubble DALK continued by injecting air in a different location distant from the perforation.

Superficial Trephination

Superficial trephination hinders the insertion of the needle at an appropriate depth. In this case deepening of the wound at the needle insertion site can be performed with a sharp knife under direct visualization (Video 5-3).

Descemet’s Membrane Perforation Before or During Air Injection

This can occur if the needle is inserted too deeply. In this case, the needle is withdrawn, an air bubble introduced in the anterior chamber through a side port to seal the perforation, and air injection is performed more superficially in a different section of the cornea, and big bubble DALK can still be completed.

If a big bubble is not obtained during the initial air injection, the injection can be repeated at a different site; usually 3 or 4 injections can be performed before complete stromal opacification that precludes visualization of the advancing needle. At this stage, anterior keratectomy is performed and air injection is performed again in the remaining deep stroma using a 30-gauge needle (Video 5-4). If the attempted reinjection fails, hydrodelamination or layer-by-layer dissection until DM is exposed can be performed. However, both techniques reduce the chance of exposing DM and increase the risk of perforation.16

In some cases where the air bubble is obtained after several attempts, the cornea may be too opaque, precluding the identification of the air bubble. One method to confirm the presence of a bubble is to drain the aqueous from the anterior chamber as the central cornea over the bubble remains firm over a soft eye. A small bubble test can also be performed.²³ This test involves injection of a small air bubble from the paracentesis. If this bubble does not migrate centrally, it indicates the presence of a big bubble occupying the central space in the anterior chamber. This is called the big bubble small bubble technique.

Type of Big Bubble Formed

Care must be taken to identify the type of bubble created. Three types of big bubbles can be achieved.¹³ A Type 1 big bubble is where the air bubble separates the pre-Descemetic layer from the deep stroma creating a large central bubble of around 8 to 9 mm in diameter. This is the preferred and more common type of bubble in DALK, as the pre-Descemetic layer confers additional strength to the recipient DM-endothelium complex. The less common Type 2 big bubble is where the DM is separated from the posterior surface of pre-Descemetic layer by the air bubble. This big bubble starts forming in the peripheral cornea toward the center, can reach the far periphery of the cornea, and has a thinner wall. Recognition of Type 2 bubble is important as it is more susceptible to tears and bursting; extreme caution is necessary during stromal dissection in these cases (Video 5-5). Type 3 is a mixed big bubble when the previous 2 coexist; usually Type 1 is complete and Type 2 is partial, but both can be complete.²⁴ For cases with mixed bubble formation, puncture of the Type 1 bubble is recommended and dissection of the stroma above the pre-Descemetic layer is performed. The Type 2 bubble is then evacuated before securing the corneal graft without further dissection at the level of bare DM. In our experience, more than 85% of the big bubbles formed during DALK surgery are Type 1. Type 2 bubbles are less common, and mixed bubble formation is rare.

The most common early postoperative complication is DM detachment. In most cases, this resolves spontaneously within the first 2 weeks. In the rare cases where a large detachment does not resolve spontaneously, single air injection in the anterior chamber usually suffices to achieve DM attachment. In the last 100 cases performed by the authors, air injection was required once (1%) to reattach DM 3 weeks postoperatively (Figure 5-6).

Stromal rejection although rare, may occur after lamellar keratoplasty (LK).^14,24,25 In our series, the incidence of stromal rejection after DALK was less than 6%. Stromal rejection usually responds well to topical steroids and resolves in most cases. Cases of stromal rejection unresponsive to medical treatment may end in corneal scarring and vascularization. In these cases a regraft is required, and it is possible to replace the full-thickness corneal stroma and preserve the recipient endothelium. This is an important advantage of DALK with exposure of DM. However, in cases with near Descemet dissection, where the DM is not exposed in the primary surgery, regrafting while preserving the DM may be difficult as the interface can be scarred and vascularized.

Stay updated, free articles. Join our Telegram channel

Join