The Continuing Evolution of Xen

Thinking outside the eye.

The Xen Glaucoma Treatment System from Allergan, which includes the XEN45 stent, was approved by the FDA in 2016 as a minimally invasive gel stent that creates a subconjunctival drainage system for the management of refractory glaucoma. Since that time, there has been tremendous evolution in surgical technique and postoperative management practice patterns for this stent. For example, many surgeons have transitioned to doing most cases under topical anesthesia, injecting undiluted mitomycin C (MMC) far posteriorly into tenons after injecting the implant, and tapering steroids as soon as all inflammation has resolved to avoid a steroid response or a “hypertensive phase.” Another new technique is ab-externo Xen implantation. To our knowledge, this technique was pioneered by Sebastian Gagne, MD, and Ike Ahmed, MD, among others. Following the presentation of surgical video by Dr. Gagne at the 2018 American Academy of Ophthalmology meeting, we began to incorporate this technique into our practice. Many in the glaucoma community are also transitioning toward the ab-externo approach for Xen implantation. The purpose of this article is to describe our current surgical technique for an ab-externo implantation of the Xen stent and also to discuss characteristics of patient selection.

Surgical Technique

We perform ab-externo Xen implantation under topical anesthesia using multiple drops of topical tetracaine ophthalmic solution. We then place a corneal traction suture to allow for control of the globe. After rotating the eye inferiorly, additional tetracaine drops are placed in the far superior forniceal conjunctiva. Using a surgical marking pen, we place ink on the needle tip of the injector, with care not to blunt the 27-gauge needle. The purpose of marking the tip is to enable location of the conjunctival incision, ensure the location is far from the tip of the Xen, confirm that it is not Seidel positive, and suture the hole if it is leaking.

Capilers are used to mark 2 mm posterior to the limbus. Nontoothed forceps are used to mobilize the superior-temporal conjunctiva around 5 mm to 6 mm posterior to the limbus and rotate it in the inferior nasal quadrant with the goal of coming within 1 mm to 2 mm posterior to the limbus (Figure 1). The stent is then injected through the conjunctiva 2 mm posterior to the limbus at or around the 12 o’clock position. Extreme care is taken to ensure visualization of the tip of the injector within the anterior chamber and that the injector is anterior and parallel to the iris. The stent should not be implanted without confirmation that the injector tip is anterior to the iris.

Figure 1. Intraoperative photograph demonstrating the use of a surgical marking pen to mark the tip of the Xen injector.

After the injector tip is confirmed to be in an ideal position, the blue sliding mechanism is slowly moved anteriorly on the handle; this will deploy the implant. When the injector needle begins to retract into the sleeve, the surgery is paused. The surgeon then holds the traction suture with the nondominant hand and allows the eye to return to its normal orthotropic position or in a position with minimal traction on the globe. The traction suture is used to steady the eye. The surgeon then continues to advance the blue slider anteriorly and completes the implantation of the stent (Figure 2). Occasionally, there can be a flick or sudden movement of the eye if there is inadvertent torque on the globe. This flick can sometimes move the implant; however, the surgeon can manipulate the Xen through the conjunctiva as when using the ab-interno approach.

Figure 2. Intraoperative photograph demonstrating the movement and dragging of the superior temporal conjunctiva in the inferior nasal direction to ensure the conjunctival button hole is not located in the position of the Xen implant. Images A, B, and C show the movement of the forceps in chronological order.

The next step is to assess the positioning of the stent. A blunt instrument or spatula is placed on the conjunctiva near the limbus and pushes the conjunctiva posteriorly. This maneuver will “milk” away any conjunctival chemosis and enhance visualization of the positioning of the stent (Figure 3). The surgeon then readjusts as needed using blunt forceps and pushes or pulls the stent anteriorly or posteriorly, through the conjunctiva, until there is roughly 1 mm of the implant in the anterior chamber, 2 mm of the implant within the sclera, and 3 mm of the implant in the subconjunctival space. A bleb should form immediately, which can occasionally obscure the view of the stent. Using a blunt instrument to milk away the chemosis will again allow visualization of the stent in the subconjunctival space.

Figure 3. Intraoperative photograph demonstrating the use of conjunctival forceps to milk away the blood and chemosis and allow for optimal visualization of the Xen implant. This image is from a different case than the case shown in Figures 1, 2, and 4.

At this point in the surgery, we will inject MMC at doses ranging from 40 µg to 80 µg, depending on the case. We aim well posteriorly to the stent and close to the fornix, but we focus on treating the area where we think aqueous will be shunted. We meticulously protect the limbus and milk MMC toward the fornix. Our goal is to inject the highest concentration and smallest volume of MMC to minimize collateral damage.

The next step is to reform the anterior chamber with intracameral preservative-free dexamethasone on a 30-gauge needle. Once the eye is pressurized, a gonioprism is used to visualize the positioning of the stent in relation to the other anterior-segment angle structures. Ideally, the stent should be parallel to the iris and just anterior to the trabecular meshwork (Figure 4). The surgeon should ensure that the stent is not too posteriorly placed. If the stent enters too posteriorly, near the iris root, it will likely become occluded by iris fibers or scar tissue and will be at high risk of failure. If the stent is too posterior in the angle or touches the iris, the surgeon should retrieve the implant via an ab-interno approach. In this situation, we prefer to then convert the surgery to an ab-interno approach; however, the surgery could theoretically be attempted with an ab-externo delivery through the same conjunctival incision, which has already been marked with surgical ink.

Figure 4. Intraoperative gonioscopic photograph demonstrating the ideal positioning of the Xen implant, anterior to the trabecular meshwork. One can appreciate a small air bubble at the tip of the implant.

Patient Selection

The following are suggested criteria for patient selection, which are not intended to be all inclusive.

Characteristics of ideal ab-externo candidates:

  • Mobile conjunctiva that can be pinched and moved without tearing;
  • Unstable intraocular lens, which would put an ab-interno approach at risk of dislocating the lens; and
  • Cloudy cornea, which makes visualization during an ab-interno approach difficult.

Characteristics of poor ab-externo candidates:

  • Thin, frail conjunctiva at risk of tearing once touched;
  • Deep-set eyes with a prominent brow, making it hard to insert superiorly; and
  • Dense corneal pannus, which makes visualization of the injector needle in the superior quadrant difficult.


Because the ab-externo innovation has not eliminated the need for performing ab-interno Xen implantation, it is an important technique to evaluate. Surprisingly, there has been a tremendous adoption of this technique without any good-quality, long-term data available on outcomes. We intend to report on our ab-externo data; however, based purely on clinical experience without having analyzed the data, there does not appear to be a dramatic difference in outcomes between the two approaches.

We feel strongly that surgeons should incorporate the ab-externo approach into their surgical armamentarium, but that it is still essential to obtain proficiency with the ab-interno method. If the ab-externo approach cannot be performed, the ab-interno method may be used alternatively. Another method is to perform Xen surgery after creating a conjunctival peritomy, and this technique has relatively good outcomes as well. If this approach is used, one can easily remove and reinject the stent until the position is perfect.

There continues to be rapid innovation in the field of surgical glaucoma and specifically with the Xen stent. Although all of these approaches need to be evaluated in a scientific manner, we continue to enjoy refining the various approaches and learning from our colleagues, who continue to innovate at a tremendous pace. GP