Though a more straightforward surgery in many respects than trabeculectomy, tube shunt procedures still have their fair share of intraoperative and postoperative challenges. Non-valved tubes in particular, require critical modifications to function at their best. But, with the right techniques, surgeons can ensure their patients receive all the pressure-lowering benefits.

Here, glaucoma surgeons share pearls and the latest techniques for getting the most out of these glaucoma drainage devices while reducing complications.


Valved or Non-Valved?

When considering the choice between a valved or non-valved tube for glaucoma surgery, surgeons weigh several key factors to tailor their approach to each patient’s needs. Eileen C. Bowden, MD, of the University of Texas at Austin, says the tube selection often comes down to the target IOP and the urgency of that pressure reduction. “Non-valved implants tend to give patients lower pressures than valved implants, so if I have a patient with more advanced disease who needs lower pressures, I generally choose a non-valved tube,” she says. “When I need the pressure down immediately, such as in cases of neovascular glaucoma, I’d be more likely to choose a valved implant.”

Surgeons say they also consider the patient’s age and natural scarring response. The likelihood of scarring is higher in a younger patient than an older patient, and younger patients may do better with non-valved tubes’ extended efficacy.

“I’ve actually moved away from implanting valved tubes because they have a lower long-term success rate compared to non-valved tubes, as shown by the ABC and AVB studies,” says Mary Qiu, MD, of the Cole Eye Institute at the Cleveland Clinic. “I try as much as possible to offer patients non-valved tubes because of this higher success rate. The trade-off is that non-valved tubes have a higher risk of hypotony.”


Why do non-valved tubes achieve lower pressures?

“The biggest influencing factor perhaps is the surface area of the plate of non-valved implants,” explains Eileen C. Bowden, MD, of the University of Texas at Austin. “They have a larger surface area. For example, one of the most common non-valved implants is the Baerveldt implant, and that comes with a 350-mm2 plate. The size of the plate ultimately determines the size of the bleb that forms once the eye has healed, and a Baerveldt’s bleb is ultimately a larger reservoir for fluid to drain into. That’s one reason we believe these non-valved implants can achieve lower eye pressures than valved implants.

“Another consideration is that the non-valved implant doesn’t have flow for the first several weeks,” she continues. “In valved implants, there’s flow immediately, and there’s thought that that immediate flow, the fluid hitting up against the Tenon’s and conjunctiva as the eye is healing, leads to the formation of a denser capsule or bleb, which limits the amount of flow that the valved implant can ultimately achieve.”


Ligating Tubes

“If we were to implant a non-valved tube and let it start functioning right away, there’d be very little resistance in the beginning and the eye pressure would drop right down to zero in most cases, leading to a host of problems,” Dr. Bowden explains. “Most surgeons have some technique of ligating the tube in the first several weeks of surgery to allow for the eye to heal and form some scar tissue at the end plate, so that when the tube does eventually start functioning when the ligature dissolves or is removed, there’s some natural resistance to outflow from the scar tissue. This allows for the eye pressure to come down gradually.”

A typical ligation consists of closing off the tube with a 7-0 Vicryl suture that’s expected to dissolve within five to six weeks. Here are some additional strategies surgeons employ to gain more control over pressure reduction:

• Irrigate the tube to ensure 100-percent ligation. “After I ligate it with a 7-0 Vicryl, I always put a 30-gauge cannula up the tube and try to irrigate it to ensure that it’s completely ligated,” says Jonathan Eisengart, MD, of the Cole Eye Institute at the Cleveland Clinic. “I’ve found especially with Ahmed ClearPath tubes, it can sometimes be difficult to get to a full ligation. You really do want to make sure that it’s completely and totally ligated so that you don’t have hypotony on postoperative day one.”

• Laser the Vicryl suture to open it earlier. If pressures are too high for too long, “using a laser can help to augment the tube function postoperatively,” says Michele C. Lim, MD, of the University of California, Davis.

• Consider using a Prolene ligature in certain cases. Dr. Bowden says she ligates with a Prolene suture instead of a Vicryl suture for certain patients, such as those who’ve had a vitrectomy in the past or who are highly myopic, who could be at risk for hypotony and subsequent suprachoroidal hemorrhage if their tube were to open before there was enough resistance to flow. “I’ll place the ligature in a location where I can see it in the clinic at the slit lamp,” she explains. “This gives me a bit more control over when the tube opens, because I can use a YAG laser to open the Prolene suture ligature, and I can do it in a more controlled fashion. I’ll do this in patients who I’m worried are at high risk of hypotony as their tube opens. Some surgeons use a Prolene ripcord suture in a similar fashion.”

• Place a ripcord suture for earlier tube functionality. When removed, a smaller diameter ripcord, such as 4-0 Prolene, can promote flow before the ligature dissolves. “You just have to be careful because if you remove the ripcord too early, before that capsule is formed enough, then you’ll get hypotony,” Dr. Eisengart cautions. “Four weeks is pushing it. Five weeks, I would say is usually pretty safe to remove a rip cord.

“Some surgeons have found that the larger 3-0 Prolene will cause some resistance to flow or some incomplete obstruction of the tube,” he continues. “The advantage of this partial obstruction is that when the ligature releases after five or six weeks, if the 3-0 Prolene is still in place it can provide some protection against hypotony. Mary Qiu, MD, does this often. I’ve used this approach occasionally in patients whom I really want that low pressure from a non-valved tube, but I’m worried about hypotony, especially early after surgery.

“Once the pressure is stabilized and you’re happy with it, you can go ahead and remove it,” he says. “In most cases, you’ll end up removing the ripcord suture eventually, but Dr. Qiu recently published an article1 describing cases where she’s left the ripcord in long-term with no apparent ill effects from it.”

“The 3-0 Prolene ripcord placed at the time of surgery partially occludes the tube lumen, so when the ligature dissolves on its own as expected at postop week six, the lumen is partly blocked by the 3-0 Prolene,” Dr. Qiu says. “This gives us a chance to titrate the pressure before removing the ripcord at a time of our choosing, if needed. For example, when the ligature dissolves, if the pressure drops too low and you have a shallow anterior chamber or choroidal effusions, then you wouldn’t want to remove the ripcord to fully open the tube. If the ligature opens and the pressure is in the mid- to high teens, and the patient is still on some baseline pressure-lowering drops, then you can remove the ripcord and stop some of the glaucoma drops.

“If the eye is inflamed when the ligature dissolves—ligature dissolution is an inherently inflammatory process—or the patient isn’t using as much steroid as instructed, then having the ripcord in place also gives you a chance to increase your steroids before choosing when to remove the ripcord,” she continues. “So, this ripcord is useful because it helps reduce or minimize hypotony-associated complications at the time of ligature dissolution, which can be a problem with non-valved tubes done without a ripcord.”

Dr. Qiu’s group published a series of patients who underwent tube shunt surgery with non-valved implants with a ripcord.2 “In two-thirds of patients, when the ligature dissolved, the pressure was good and the eye wasn’t inflamed. We were able to remove the ripcord at the next visit after the tube opened and stop some glaucoma drops. Those eyes did great. In the remaining one-third of patients, some patients’ pressures were already quite low and we were able to stop some drops before removing the ripcord later; some eyes were inflamed, and we were able to increase steroids and remove the ripcord later; and one eye with persistent single-digit IOP had the ripcord trimmed and left in place permanently. There were no hypotony-associated or infectious complications.

“For patients [such as the latter] at high risk for hypotony (e.g., from uveitis, neovascular glaucoma or prior cyclodestructive laser procedures), you could plan to leave the ripcord in for longer than you would otherwise, or indefinitely,” she adds. “The ripcord can be trimmed so it’s entirely under the conjunctiva with nothing exposed, and you can always cut down and retrieve it later.” 

While waiting for non-valved tubes to become functional, surgeons say they’ll usually continue the patient’s glaucoma medications. Additionally, some may “start or continue oral acetazolamide, a carbonic anhydrase inhibitor that helps make a big impact in lowering the eye pressure during the period the tube is ligated,” Dr. Bowden says.

Figure 1. A right eye, showing the tube in position from the surgeon’s view. The plate is sutured into the superotemporal quadrant. The tube enters the anterior chamber near 12:00 through a 3-mm-long scleral tunnel. Dr. Eisengart believes this is optimal for reducing the risk of tube erosion. A scleral reinforcement graft is covering part of the cornea and will be sewn in place next. All images: Jonathan Eisengart, MD.


The use of fenestrations is surgeon- and patient-dependent, says Dr. Bowden. “If you have a patient who can’t tolerate oral acetazolamide as we wait for the tube to open, for example, I might make two fenestrations in the implant so there’s some flow,” she says. “In most cases, however, just because I find fenestrations can be unpredictable sometimes depending on the needle you use or where you place your fenestrations, I try to reserve fenestrations for when I really need them and usually in patients who can’t tolerate their glaucoma medications.”

Using the needle of the 7-0 Vicryl to pierce the tube a few times distal to the ligature is a common way of making fenestrations. Dr. Eisengart says he used to make his fenestrations this way but was finding the pressures still higher than he wanted them to be. “So, for a while I switched to using a 15-degree blade and making a larger fenestration in the tube. I did that for a couple years, and I found I was getting lower pressures but also that there was no way to standardize the size of these [fenestrations]. The fenestration you make with a 15-degree blade varies depending on how far through the tube you stick the triangular blade.”

Now, when implanting non-valved tubes, he’s using a single 10-0 nylon wick suture. “It seems to work really well for giving me good pressure control early in the postoperative period without causing too much hypotony,” he says. “I take a 10-0 nylon on the needle and pass it straight through the tube, and then cut the needle off. I leave about a five- or six-millimeter piece of nylon right through the tube. In my hands, that seems to produce just enough flow to control the pressure.” 

Wick sutures or the “vent and stent” technique, Dr. Lim says, “allows aqueous humor to trickle out. It has a much more profound effect on fluid exit than doing a fenestration where you don’t leave a length of suture in passing through the tube. So, this can give you some good control for several weeks after surgery until the tube opens on its own.” 


Adding Goniotomy

“A couple of years ago, I started doing a technique where I add a goniotomy to my non-valved tubes in eyes that have an open angle; that aren’t on aspirin or blood thinners; that haven’t had prior angle surgery; and don’t have PAS in the angle,” Dr. Qiu says. “This way, the goniotomy can help provide some early pressure lowering and the aqueous can flow out the goniotomy site. You don’t get as much hypotony because of the episcleral venous pressure. This way, the pressure can be lowered immediately, even with a non-valved tube, and without the same risk of hypotony-associated complications that can occur if fenestrations or the wick are too big.

“We recently published our series3 of non-valved tubes combined with goniotomy,” she continues. “My early results show that patients achieve good pressure before the ligature dissolves, and no one had any hypotony-associated complications. We presented our one-year outcomes in a poster4 at the American Glaucoma Society meeting earlier this year.”


Supramid to the Rescue 

Despite surgeons’ best efforts to prevent hypotony early on and when the ligature dissolves, late hypotony with non-valved tubes can strike. “If the eye’s outflow is greater than the aqueous production, the pressure is going to be too low,” Dr. Qiu says. “In eyes that have a non-valved tube, down the line, they may end up with too-low pressures and the potential for hypotony-associated complications such as hypotony maculopathy or choroidal effusions that don’t improve with medical management. You can always go back and revise the tube to decrease the amount of flow exiting the non-valved tube. 

“Historically, this has been done by adding a ligature onto the tube either by opening the conjunctiva or by ligating the tube inside the anterior chamber,” she continues. “More recently, there have been techniques describing the insertions of a stent suture into the lumen of the tube ab interno through multiple paracenteses. My latest strategy to address late hypotony in a non-valved tube is to insert a segment of 3-0 Supramid suture into the tube lumen through a single paracentesis in an ab interno approach. If the ripcord has been removed, you can put a very long piece of 3-0 Supramid up the tube, and it will go all the way up into the capsule. This will block flow—3-0 Supramid is a twisted suture, so it expands a bit when hydrated, providing some occlusion of the tube lumen. Even if the original ripcord hasn’t been removed, you can still put a shorter piece of 3-0 Supramid up the tube, just up until it reaches the position of the ripcord. 


Obstruction & Fibrosis

Tube shunts can become blocked by materials such as iris pigment or heme. “Much of the time, if you wait a few days, [the blockage] will flush through and clear out,” Dr. Bowden says. “I’ve tried using a YAG laser to dissolve the blockage if I’m able to easily visualize it. But, if it’s been several days and the blockage is still in place, I’ll take the patient back to the operating room, remove the tube from the anterior chamber and flush it with saline to clear the blockage manually.”

According to Dr. Lim, tube blockage is less of an issue than when the bleb over the end plate becomes thickened due to fibrosis. “In the hypertensive phase, about two to three months after surgery, patients tend to experience a pressure rise, which we think is the bleb remodeling over the implant plate,” she explains. “Some surgeons will inject more anti-metabolite medication around the tube area in the postop period if they think there’s a lot of scarring. Right now, there’s not a lot of good evidence on the outcomes of doing that. Some studies are coming out of University of California in San Francisco, looking at the outcomes of injecting mitomycin-C at the time of surgery. Their studies are showing that it can augment the outcome of these implants.”

“Even after revision, there’s a chance the fibrosis may form again,” Dr. Bowden says. “It’s more common in the valved implants but can occur in non-valved implants as well. Some surgeons advocate for using anti-fibrotic materials like mitomycin-C or 5-flurouracil but there haven’t been any large studies or randomized controlled trials showing the best ways to do this. Some advocate injecting mitomycin-C or 5-FU when you place the tube and others advocate for injecting that into the subconjunctival space the week after the tube is placed, while the eye is still inflamed.

“Others use a needle several months after a tube is placed if there’s significant fibrosis,” she continues. “Using a needle as well as an anti-fibrotic to manually break up the scar tissue over the plate can be done either at the slit lamp in the office or back in the operating room. I’ve found that whenever there’s a lot of fibrosis over the plate, it’s best to take the patient to the operating room and take down the conjunctiva and sub-Tenon’s and then remove the capsule manually, cut it out and close the conjunctiva again.”

Figure 2. An eroded Ahmed tube is seen in the upper right of this video screen capture. The Ahmed plate is visible as the capsule overlying it is being excised. In this capture, the capsule overlying the plate has been cut on three sides, and the final cut along the posterior edge will free the graft. Not all patients are good candidates for this approach because they must have sufficient capsule that can be safely excised.

Sulcus Placement

“One of the downsides of tube shunts versus trabeculectomies is that there’s a foreign body in the eye,” Dr. Qiu says. “Anterior chamber placement in particular can increase the risk of corneal decompensation, and then these eyes may require corneal transplants such as endothelial keratoplasty.”

To reduce the risk of corneal decompensation, Dr. Qiu preferentially places tubes in the ciliary sulcus whenever possible when the eye is pseudophakic. “I have a relatively low threshold to do cataract surgery concurrently in eyes that need tubes if they qualify for cataract surgery, so that I can make the eyes pseudophakic and place my tube in the sulcus instead of in the anterior chamber. We need more long-term data on this, but right now, the data show that there’s less endothelial cell loss with sulcus tubes compared to anterior chamber tubes.”

Sulcus placement can be challenging, for a number of reasons. “You’re inserting the needle behind the iris, and you can’t always see exactly where it’s going to come out,” Dr. Qiu says. “Sometimes the needle enters the vitreous instead inadvertently or pokes into the iris root instead of the sulcus. Sometimes the needle does successfully enter the sulcus where you want it, but when you try to insert the tube into the needle track, it doesn’t want to go, because unlike the rigid, metal needle, the tube is floppy. The trade-off with sulcus placement is that sometimes the sclerotomy is quite posterior and the tubes may end up behind the IOL.

“To place tubes in the sulcus more predictably, we’ve recently been using a guidewire,” Dr. Qiu continues. “There are multiple ways to do this, and we recently presented three different approaches at AGS.5 Videos demonstrating how to do this are also published in AJO Case Reports.6 My approach of doing guidewire-assisted sulcus tube entry is to first inflate the eye with viscoelastic. I prefer Healon. Make a paracentesis 180 degrees across the eye from where the tube entry site is going to be. Have a segment of 3-0 Prolene suture cut and ready. (This is the same 3-0 Prolene suture I use for my ripcord, so it’s already open and on the field.) Once the needle entry is made into the sulcus, the guidewire can be inserted into the paracentesis with the other hand and docked into the bevel of the needle and then pulled through the sclerotomy site. Now there’s a segment of 3-0 Prolene in the sulcus for the tube to follow. Put the tube onto the guidewire and push it into the sulcus. The tube will follow the guidewire into the sulcus, and it can’t end up behind the IOL. I like my sulcus tubes to sit as posteriorly as possible, so they’re far away from the iris.”


Tube Erosions

Another potential challenge of tube shunts is that they may erode or become exposed, opening pathways for infection such as tube-associated endophthalmitis. Tube erosions occur in two main places: 1) over the tube path and 2) over the plate. “It’s far less common for the conjunctival tissue to break down over the plate of the tube itself, but when it happens these types of erosions are very hard to fix,” Dr. Lim says. “In fact, when we see these erosions, we often think that we won’t be able to fix it and get it to work long-term.”

In a poster presentation at AGS this year, Dr. Lim shared retrospective data on outcomes of tube erosion repair.7 “We were interested in the rate of re-erosion based on where the initial erosion occurred,” she says. “Our study showed that the recurrence rate of erosion, when the erosion happened over the tube, was much lower than when it happened over the plate. The overall sample size was small, however, so you have to take this study with a grain of salt.”

Dr. Bowden says that many of the tube erosions that she’s seen over time have been due to an anteriorly placed tube. “I try to place the tube at least eight millimeters posterior to the limbus to minimize the risk of erosions,” she says. “I also like to use a patch graft of some sort, whether that’s sclera or cornea, to cover the tube adjacent to the limbus.”

 “For cases where you’re placing a tube in the anterior chamber, which is probably where a majority of tubes still go, that means creating a two- to three-, but ideally three-millimeter scleral tunnel,” Dr. Eisengart adds (Figure 1). “So rather than just putting the tube directly into the anterior chamber at the limbus, you want to start more posteriorly, about three millimeters back, and tunnel the tube through sclera. In my experience this really reduces your erosion rate. The downside of tunneling is that it’s a little more challenging to get the angle right and to try to get the entry parallel to the iris and posterior enough.”

To reduce the risk of erosion due to eyelid rubbing, he recommends positioning the tube entry site close to the 12 o’clock position. “I find that a lot of erosions happen where the lid margin rubs over the tube. Generally, that’s going to be where the lid margin crosses the limbus superotemporally and superonasally, at roughly the 10:30 and 1:30 positions. So, when I’m putting a tube in the anterior chamber, I try to make the entry site closer to the 12 o’clock position. I think we have fewer erosions if we can get the tube to enter around 12 o’clock rather than entering superotemporally or, much less commonly, superonasally. 

If an erosion occurs, there are two key things to consider:

1. Change the location of the tube. “You may want to think about repositioning the tube,” Dr. Eisengart says. “The easiest thing to do is to simply re-cover the tube where it is, but if it eroded in that area once, there may be a risk of recurrent erosion if you just recover it.”

2. Switch graft materials. In the same vein, experts recommend using a different patch graft material for the repair. “Donor sclera is the most common way we [cover when] we put in tubes, so the surgeon might use donor cornea instead because it tends to be more durable than sclera,” Dr. Eisengart says.

“Tutoplast, which is preserved human pericardial tissue, is another good option,” Dr. Lim adds. “I haven’t seen any good studies comparing different patch graft materials but there’s some thinking that corneal tissue would have greater coverage longevity because the collagen fibers of the cornea are a lot tougher, perhaps, than what you’d find in an engineered Tutoplast or a scleral patch graft.

Figure 3. Here, the anterior conjunctiva in the area of the tube erosion has been dissected off of the tube and up to the limbus and reflected anteriorly. The harvested capsule autograft has been placed overlying the tube. The actual conjunctival erosion hole is visible toward the upper right (between the graft and the forceps).

Using the Capsule 

When faced with a tube erosion, it can sometimes be difficult to mobilize enough tissue to cover the tube. If that’s the case, or the surgeon isn’t sure which type of patch graft material they may want to use, Dr. Qiu suggests using some of the capsule itself as graft material. 

“For Ahmed erosions (a valved tube), you can harvest a piece of the capsule that’s grown around the Ahmed end plate and use that autograft to patch up the tube erosion (Figures 2 and 3). This autograft is autologous and vascularized, so it may have some advantages over other patch graft materials such as irradiated donor sclera, donor pericardium or donor cornea that’s not autologous to the patient. We think that this capsule autograft may integrate into the body’s tissues a little better. 

“As with other types of patch grafts, you can leave the autograft uncovered by conjunctiva because conjunctiva will grow over it,” she continues. “So, if you’re able to harvest a piece of the Ahmed capsule and put that over the part of the tube that’s eroded, you can then cover it up with conjunctiva as best as you can. But even if it’s not fully covered, conjunctiva will grow over it. This autograft material is nice because it’s already very spongy and vascular.”

When attempting this capsule autograft approach with a non-valved tube, Dr. Qiu cautions that “you can’t simply cut off a piece of the capsule, use it as a patch graft and walk away.” She explains, “If you violate the capsule [of a non-valved tube] you’ll get hypotony. So, when doing this with a non-valved tube, you have to re-ligate the tube so that the body has time to regrow the capsule around the end plate. You can still harvest the capsule autograft; you just have to re-ligate the tube.

Figure 4. The eye 17 months after the repair using a capsule autograft. Capsule autografts have good cosmetic results.

“Dr. Kahook has published a case series using this technique for Ahmed erosions, specifically, and my subsequent case series demonstrates the use of this capsule autograft for a variety of indications,” she says, noting that it’s useful when placing a second tube or performing a tube exchange in addition to tube erosion. “I’ve now done this technique in more than 20 eyes, and we’re working on publishing an updated case series with photos of everyone’s outcomes. The capsule autograft has a more cosmetic result compared to other types of patch grafts (Figure 4). Scleral patch grafts, for example, look like little white rectangles sitting on the surface of the eye, whereas the autologous capsule patch graft just looks like the body’s own tissue. I’ve only been doing this technique for the last four years, so we don’t have long-term data about how well these patch grafts hold up in terms of erosion rates. I’m looking forward to longer term outcomes on these patients.” 

Drs. Bowden, Lim, Eisengart and Qiu have no related financial interests.



1. Wang J, Chun LY, Qiu M. Baerveldt-350 with 3-0 Prolene ripcord to minimize hypotony-associated complications after spontaneous ligature dissolution. Ophthalmol Glaucoma 2024; 7:1:93-100.

2. Chun LY, Qiu M. Utility of 3-0 Prolene ripcord with Baerveldt-350 to partially occlude tube lumen and minimize hypotony-associated complications when ligature dissolves. Invest Ophthalmol Vis Sci 2022;63:3693-A0378.

3. Kanter J, Qiu M. Baerveldt-350 with adjunctive goniotomy: Pilot results. Am J Ophthalmol Case Rep 2023;32:101950.

4. Kanter J, Qiu M. Baerveldt-350 with adjunctive goniotomy: One-year outcomes. Poster presented at the 2024 American Glaucoma Society Meeting in Huntington Beach, Calif.

5. Qiu M. 3 Techniques for guide-wire assisted sulcus tube entry. Surgical Video Series. Presented at the 2024 American Glaucoma Society Meeting in Huntington Beach, Calif.

6. Shah A, Kanter J, Eisengart J, Blieden LS, Qiu M. Three techniques for guidewire-assisted sulcus glaucoma tube shunt placement. Am J Ophthalmol Case Rep 2024;34:102009.

7. Jeong S, Gale M, Jundi M, Lim MC. Location of glaucoma drainage device erosion and relationship to success of repair. Poster presented at the 2024 American Glaucoma Society Meeting in Huntington Beach, Calif.