תעתיק 🔗
WEBVTT
00:00:00.000 --> 00:00:04.719
A steroid injection during glaucoma surgery was linked to dangerous low eye pressure.
00:00:04.879 --> 00:00:07.200
What a new randomized trial found.
00:00:07.440 --> 00:00:10.720
Research checked through November 4th, 2026.
00:00:11.199 --> 00:00:17.839
A small randomized trial published on October 29, 2026, raised a serious safety question.
00:00:18.079 --> 00:00:26.079
Could a long-acting steroid injection given during glaucoma drainage device surgery suppress healing so much that the eye drains too freely?
00:00:26.320 --> 00:00:36.640
In the trial, late-onset low eye pressure occurred in 10 of 19 eyes given triamcinolone around the implant, compared with none of 20 control eyes.
00:00:36.880 --> 00:00:40.399
Seven of the 10 affected eyes needed another operation.
00:00:40.560 --> 00:00:52.079
That is a striking signal, but it does not show that ordinary steroid eye drops after glaucoma surgery are dangerous, nor that every drainage device or steroid regimen carries the same risk.
00:00:52.479 --> 00:01:01.600
The study tested one 20 milligram depot injection with one non-valved implant in a small group of people with difficult-to-control glaucoma.
00:01:01.920 --> 00:01:11.200
Important details about the pressure threshold, the timing of individual cases, and their recovery are not reported in the publicly available abstract or registry.
00:01:11.439 --> 00:01:20.079
Trial report on PubMed, trial record on Rebecca, what the new randomized trial tested, treatment, injection location, and masking.
00:01:20.719 --> 00:01:26.400
The study included 39 eyes from 39 people with refractory primary open angle glaucoma.
00:01:26.560 --> 00:01:30.239
All were undergoing Susanna UF glaucoma drainage device surgery.
00:01:30.400 --> 00:01:34.959
The registry says participants had failed at least one prior trabeculectomy.
00:01:35.200 --> 00:01:40.879
20 eyes were assigned to the device alone, and 19 to the device plus triamcinolone.
00:01:41.040 --> 00:01:42.719
Follow-up lasted six months.
00:01:42.959 --> 00:01:46.560
The journal describes the study as multi-center and randomized.
00:01:46.719 --> 00:01:50.000
Trial report on PubMed trial record on Rebecks.
00:01:50.400 --> 00:01:57.519
The triamcinolone was 20 milligrams, 0.5 milliliters of a 40 mg per milliliter suspension.
00:01:57.760 --> 00:02:06.400
It was injected into the subtenon space around the implant plate immediately before the conjunctiva, the tissue covering the white of the eye, was closed.
00:02:06.560 --> 00:02:12.000
This is a depot injection beside the implant, not an injection into the inside of the eye.
00:02:12.319 --> 00:02:14.719
Trial record on Rebecca.
00:02:15.039 --> 00:02:19.039
The published abstract says that patients and outcome assessors were masked.
00:02:19.199 --> 00:02:25.280
The registry says a sham injection might have been used to mask the surgeon, but also allows that no sham was performed.
00:02:25.439 --> 00:02:30.159
So
Load full transcript
based on the public records, surgeon masking cannot be confirmed.
00:02:30.639 --> 00:02:34.000
The registry also does not provide the full surgical recipe.
00:02:34.159 --> 00:02:41.120
It does not establish whether the tube was tied, fenestrated, or internally stented, or whether mitomycin C was used.
00:02:41.280 --> 00:02:45.520
Those details matter because they can change how much fluid flows through the device.
00:02:45.759 --> 00:02:48.960
Trial report on PubMed, trial record on Rebecca.
00:02:49.120 --> 00:02:50.560
There are two further cautions.
00:02:50.719 --> 00:02:56.159
The registry lists a target of 48 participants, but only 39 were enrolled.
00:02:56.400 --> 00:02:59.919
Recruitment stopped early because of safety concerns.
00:03:00.240 --> 00:03:07.759
It also records a first enrollment date in December 2020 and a public registration date in July 2025.
00:03:08.159 --> 00:03:18.800
Late public registration does not prove that results were reported selectively, but it limits confidence that every outcome and analysis was publicly specified before enrollment began.
00:03:19.039 --> 00:03:27.199
All 39 enrolled participants completed six-month follow-up, trial record on Rebecca, what the trial found.
00:03:27.520 --> 00:03:31.520
For the full table, please open this article on VisualFieldTest.com.
00:03:31.759 --> 00:03:36.240
The injection did not provide an additional pressure-lowering benefit at six months.
00:03:36.400 --> 00:03:43.199
Average pressure was about 14 millimeter her grams in both groups, and the reduction from baseline was almost identical.
00:03:43.439 --> 00:03:48.400
A similar group average does not rule out serious low pressure episodes in individual eyes.
00:03:48.560 --> 00:03:51.199
Averages can hide a wide range of outcomes.
00:03:51.520 --> 00:03:53.360
Trial report on PubMed.
00:03:54.800 --> 00:03:59.919
The seven operations were performed in seven of the ten eyes with late onset hypotenuse.
00:04:00.080 --> 00:04:08.159
The public report does not say what procedures were used, when each operation occurred, or whether each eye's pressure and vision ultimately recovered.
00:04:08.400 --> 00:04:10.000
Trial record on Rebecca.
00:04:10.159 --> 00:04:12.960
What is and is not known about the complications, BE?
00:04:13.520 --> 00:04:16.079
The registry reports these additional events.
00:04:16.319 --> 00:04:20.480
For the full table, please open this article on VisualFieldtest.com.
00:04:20.720 --> 00:04:25.600
The registry defines a serious complication group as late hypotony or malignant glaucoma.
00:04:25.680 --> 00:04:31.040
That composite was 52.6% in the triumphalone group and 5% in controls.
00:04:31.199 --> 00:04:33.839
The single control event was malignant glaucoma.
00:04:34.079 --> 00:04:35.920
It was not a case of late hypotony.
00:04:36.160 --> 00:04:47.120
The published abstract separately reports late hypotony as 10 of 19 versus 0, and clinically significant hypotany requiring reintervention
as 7 of 19 versus 0.
00:04:47.360 --> 00:04:49.120
Trial report on PubMed.
00:04:49.360 --> 00:04:50.959
Trial record on Rebec.
00:04:51.360 --> 00:04:59.839
The public abstract and registry do not give an exact numerical definition of hypotony or the threshold used to label an event late onset.
00:05:00.079 --> 00:05:03.920
They also do not give the month each case began or how long it lasted.
00:05:04.079 --> 00:05:14.639
The registry reports no statistically significant difference in average visual acuity at follow-up, and no event directly related to the intervention led to loss of light perception.
00:05:14.879 --> 00:05:27.040
But it does not report how many individual patients lost lines of vision or provide separate counts for choroidal detachment, hypotany-related macular changes, shallow front chamber, or bleb leaks.
00:05:27.360 --> 00:05:32.319
Those outcomes should be treated as not reported, not as confirmed zero events.
00:05:32.560 --> 00:05:34.160
Trial record on Rebecca.
00:05:34.319 --> 00:05:43.360
The registry says that de-identified participant-level data may be requested from the investigators, subject to a research proposal and ethics approval.
00:05:43.600 --> 00:05:48.800
That data set would be needed to learn the exact case timelines and individual outcomes.
00:05:49.120 --> 00:05:50.720
Trial record on Rebecca.
00:05:50.879 --> 00:05:54.560
Why might a steroid depot lead to too much drainage?
00:05:54.800 --> 00:05:57.839
The capsule around an implant helps regulate flow.
00:05:58.000 --> 00:06:01.759
A drainage device carries fluid from inside the eye to tissue around its plate.
00:06:01.920 --> 00:06:06.000
As healing proceeds, a fibrous capsule forms around that plate.
00:06:06.240 --> 00:06:12.800
The capsule is not merely scar tissue to eliminate, it also provides much of the resistance that regulates drainage.
00:06:13.040 --> 00:06:16.480
Too much scarring can restrict flow and raise eye pressure.
00:06:16.720 --> 00:06:21.199
Too little capsule resistance can allow excessive flow and lower pressure too far.
00:06:21.439 --> 00:06:25.439
American Academy of Ophthalmology overview of drainage devices.
00:06:25.680 --> 00:06:30.800
Evidence supports the general pathway, but not the precise explanation for this trial.
00:06:31.040 --> 00:06:39.040
A broad review describes how inflammation and fibroblast activity contribute to wound healing and scarring after glaucoma surgery.
00:06:39.279 --> 00:06:48.879
A separate report described three patients on long-term systemic immune suppressing treatment who developed persistent hypotany after non-valved implants.
00:06:49.040 --> 00:06:51.759
Their implant capsules were thin and immature.
00:06:52.079 --> 00:06:58.240
That makes it biologically plausible that strong suppression of healing could reduce capsule resistance.
00:06:58.399 --> 00:07:06.399
But those cases involved chronic systemic treatment, not a one-time 20 mg injection of triamzinolone around a Susanna device.
00:07:06.560 --> 00:07:10.639
The Susanna trial did not measure capsule thickness or fluid flow.
00:07:11.040 --> 00:07:12.399
Wound Healing Review.
00:07:12.639 --> 00:07:14.720
Capsule Histology Case Report.
00:07:14.879 --> 00:07:17.920
The most plausible explanation is therefore an inference.
00:07:18.160 --> 00:07:24.800
If the depot reduced inflammation and capsule formation around the plate, the outflow pathway may have had less resistance.
00:07:24.959 --> 00:07:31.600
That could lead to excess drainage, especially once any temporary restriction on tube flow weakens or opens.
00:07:31.839 --> 00:07:40.319
The trial's public records do not establish the timing of hypotany relative to tube opening or even the exact tube restriction method used.
00:07:40.480 --> 00:07:44.160
So this mechanism is reasonable but unproven for these eyes.
00:07:44.399 --> 00:07:47.600
A depot is not the same as adjustable eye drops.
00:07:47.839 --> 00:07:52.720
A depot injection leaves a suspension near the implant that releases medicine over time.
00:07:52.879 --> 00:08:00.560
A study of subtenin' triamsinolone after cataract surgery describes drug delivery lasting several weeks, though the duration can vary.
00:08:00.720 --> 00:08:03.839
In contrast, eye drops deliver a dose with each use.
00:08:04.079 --> 00:08:09.040
Clinicians can taper or stop them based on the eye's inflammation and healing.
00:08:09.360 --> 00:08:15.040
Randomized study of 20 milligram subtenon's triamsinolone after cataract surgery.
00:08:15.199 --> 00:08:25.680
Crucially, both arms in the Susanna trial received postoperative dexamethasone eye drops, initially six times a day, then gradually reduced over six to eight weeks.
00:08:25.920 --> 00:08:29.279
The randomized difference was the added depote injection.
00:08:29.519 --> 00:08:35.200
The trial therefore does not show that routine postoperative steroid drops caused the hypotenuse signal.
00:08:35.360 --> 00:08:40.799
It tested a long-acting injection on top of the same drop regimen given to controls.
00:08:41.440 --> 00:08:48.480
Trial record on Rebecca steroids can also raise eye pressure in some patients, particularly with prolonged exposure.
00:08:48.720 --> 00:08:52.080
That is a different effect from the suspected mechanism here.
00:08:52.320 --> 00:09:02.080
The concern is not that triumphinolone directly lowers pressure inside the eye, but that it might change wound healing and reduce resistance around a drainage pathway.
00:09:02.320 --> 00:09:05.279
How the findings compare with other glaucoma surgeries.
00:09:05.440 --> 00:09:09.679
For the full table, please open this article on VisualFieldtest.com.
00:09:10.159 --> 00:09:11.759
Why the results may differ.
00:09:12.000 --> 00:09:14.159
Valved versus nonvalved devices.
00:09:14.320 --> 00:09:18.080
The Ahmed valve is designed to restrict flow as pressure becomes low.
00:09:18.320 --> 00:09:23.279
Susanna UF and Bearvelt implants are non-valved, so they lack that
same built-in valve.
00:09:23.440 --> 00:09:34.240
Nonvalved implants are commonly managed with temporary tube ligatures, stents, or other flow control steps, but the Susanna trial's public records do not say exactly which were used.
00:09:34.399 --> 00:09:50.480
In a pooled analysis of three randomized tube surgery trials, chronic hypotany was uncommon overall, 4.1% by five years, but the risk was higher with Bearvelt than with AMED and was also associated with uveitic glaucoma and poorer preoperative vision.
00:09:50.639 --> 00:09:53.519
That analysis was not a test of triamcinolone.
00:09:53.679 --> 00:09:57.440
Pooled analysis of chronic hypotany, tube resistance, and geometry.
00:09:57.519 --> 00:10:04.320
The Susanna device is non-valved and has a reported tube diameter of about 230 micrometers internally.
00:10:04.879 --> 00:10:13.279
Proserflow is also non-valved, but it has a much narrower, approximately 70 micrometer lumen designed to add flow resistance.
00:10:13.600 --> 00:10:21.759
Laboratory testing found that this resistance alone may not prevent hypotany unless resistance from surrounding tissue also helps control flow.
00:10:22.000 --> 00:10:26.480
Thus, non-valve does not mean that all devices have the same flow characteristics.
00:10:26.720 --> 00:10:31.600
Susanna device description per Surflow Flow Study, Dose and Treatment Location.
00:10:31.840 --> 00:10:34.159
The Susanna trial used 20 mg.
00:10:34.799 --> 00:10:42.080
The 2017 AMED randomized trial used 10 mg, and the 2022 trabeculectomy trial used 4 mg.
00:10:43.120 --> 00:10:50.559
The Susanna and Ahmed studies both place triamcinolone near the implant plate, but they use different devices and populations.
00:10:51.120 --> 00:10:56.480
These comparisons raise a dose question, but they do not establish a dose response relationship.
00:10:56.639 --> 00:10:59.600
The studies were not designed to compare doses directly.
00:10:59.759 --> 00:11:01.120
Mitomycin C.
00:11:01.360 --> 00:11:09.759
The October 2026 Preserflow study explicitly combined triamcinolone with mitomycin C, an anti-scarring medicine.
00:11:10.000 --> 00:11:17.600
Public records for the Susanna randomized trial do not specify whether mitomycin C was used, so the studies cannot be compared on that basis.
00:11:17.840 --> 00:11:28.960
In trabeculectomy studies, adding triamcinolone to mitomycin C has not produced a consistent hypoteny finding, but those procedures and doses differ from the Susanna operation.
00:11:29.679 --> 00:11:32.000
Patient group and outcome definitions.
00:11:32.240 --> 00:11:39.039
The Susanna trial enrolled people with refractory primary open angle glaucoma and excluded uveitic glaucoma.
00:11:39.360 --> 00:11:43.039
The AMED trials included different types of refractory glaucoma.
00:11:43.200 --> 00:11:46.960
The trabeculectomy trials also differed in disease type and follow-up.
00:11:47.120 --> 00:11:49.759
Studies may also define low pressure differently.
00:11:49.919 --> 00:11:58.639
The Susanna trial's numerical threshold and exact time definition for late onset hypotany are not public in the abstract or registry.
00:11:58.799 --> 00:12:03.200
These differences make simple percentage-to-percentage comparisons risky.
00:12:03.679 --> 00:12:05.600
Too much healing versus too little.
00:12:05.759 --> 00:12:08.879
Glaucoma surgery depends on a controlled amount of healing.
00:12:09.120 --> 00:12:16.799
If the eye heals too aggressively, scar tissue can narrow or close the drainage route, causing high eye pressure and surgical failure.
00:12:16.960 --> 00:12:23.519
If healing is suppressed too much or flow is not otherwise adequately restricted, the pathway may drain excessively.
00:12:23.679 --> 00:12:33.360
The result can be hypotany, a very shallow front chamber, choroidal complications, macular folds, reduced vision, or the need for surgical correction.
00:12:33.600 --> 00:12:38.000
Wound Healing Review, American Academy of Ophthalmology review of hypotony.
00:12:38.399 --> 00:12:41.440
The trial does not identify an optimal amount of healing.
00:12:41.679 --> 00:12:48.720
It tested one 20 mg depot against no depot, while both groups received the same postoperative drops.
00:12:49.039 --> 00:12:54.639
It did not compare several doses, image the implant capsule, or measure fluid resistance.
00:12:54.799 --> 00:13:01.039
It therefore suggests that this particular regimen may oversuppress healing in this particular setting.
00:13:01.200 --> 00:13:06.080
It does not tell surgeons what the ideal capsule or scar tissue should look like.
00:13:06.320 --> 00:13:09.519
Different wound control approaches have different risks.
00:13:09.759 --> 00:13:14.399
Steroids reduce inflammation and are routinely used after filtration surgery.
00:13:14.559 --> 00:13:18.399
A depot may be harder to adjust than drops once injected.
00:13:18.799 --> 00:13:23.120
Mitomycin C and five fluorouricil limit cell growth and scarring.
00:13:23.279 --> 00:13:30.799
They can improve drainage success, but excessive suppression can contribute to fragile or leaking filtering tissue and low pressure.
00:13:31.120 --> 00:13:41.600
Cochrane review of mitomycin C versus five fluorobevicismab, which blocks vascular endothelial growth factor, has been studied as a way to reduce blood vessel growth and scarring.
00:13:41.840 --> 00:13:43.360
Results are mixed.
00:13:43.600 --> 00:13:47.840
It has not displaced established approaches as a clearly superior option.
00:13:48.000 --> 00:13:59.759
Review of glaucoma wound modulation flow control techniques, such as a valve, ligature, or tube stent, change fluid movement mechanically rather than relying only on the healing response.
00:14:00.000 --> 00:14:10.320
They are a plausible way to match flow to a device's design, but no trial has yet shown that tailoring steroid treatment to each device reliably improves both safety and success.
00:14:10.559 --> 00:14:12.480
Who might be at higher risk?
00:14:12.799 -->
00:14:16.879
The Susanna trial is too small to identify reliable predictors.
00:14:17.039 --> 00:14:20.559
The groups had similar average age and pre-operative pressure.
00:14:20.720 --> 00:14:25.120
Participants had refractory glaucoma and prior trabeculectomy failure.
00:14:25.360 --> 00:14:36.879
The public report does not provide enough information to assess myopia, scleral thickness, individual inflammatory status, or whether any particular tube control method predicted hypotony.
00:14:37.200 --> 00:14:39.039
UVatic glaucoma was excluded.
00:14:39.200 --> 00:14:47.600
Trial record on Rebecca evidence from other settings points to several possible risk factors, but they should not be mistaken for findings from this trial.
00:14:47.840 --> 00:14:57.200
In a pooled analysis of tube surgery trials, chronic hypotony was associated with barevelt implantation, uveatic glaucoma, and poorer preoperative vision.
00:14:57.360 --> 00:15:06.960
Studies of trabeculectomy have also linked prolonged hypotony or hypotony-related macular changes with factors such as younger age and myopia.
00:15:07.279 --> 00:15:15.200
These findings may help guide future research, but they do not establish who would be at risk from a triamolone depot with a Susanna device.
00:15:15.519 --> 00:15:19.519
Pooled tube surgery analysis, Trabeculectomy hypotenuse.
00:15:20.080 --> 00:15:22.080
How uncertain is the safety signal?
00:15:22.240 --> 00:15:28.480
In the Susanna trial, late hypoteny affected 10 of 19 treated eyes and 0 of 20 controls.
00:15:28.960 --> 00:15:35.360
Using Wilson confidence intervals, the estimated rate in treated eyes is roughly 32% to 73%.
00:15:35.759 --> 00:15:41.440
For zero events in 20 controls, the data are still compatible with a rate up to about 16%.
00:15:42.080 --> 00:15:48.559
A Newcomb-style estimate for the difference in rates is approximately 26 to 73 percentage points.
00:15:48.799 --> 00:15:51.679
These wide ranges reflect the small sample.
00:15:52.080 --> 00:15:57.600
Trial report on PubMed The event pattern is unlikely to be a chance imbalance alone.
00:15:57.840 --> 00:16:09.519
Under a simple model comparing the two event rates with independent, uniform prior assumptions, the probability that the underlying late hypotenuse rate was higher with the injection is well over 99.9%.
00:16:10.240 --> 00:16:21.120
But that is a probability about the rate difference in a study-like setting, not a probability that 52.6% of patients will be harmed at every center or with every device.
00:16:21.679 --> 00:16:25.440
The randomized comparisons supports a real safety signal.
00:16:25.679 --> 00:16:28.799
The small sample and early stop make its size uncertain.
00:16:28.960 --> 00:16:38.080
A simple model that shrinks the observed rates toward the middle estimates and excess risk of roughly 48 percentage points for this exact regimen.
00:16:38.240 --> 00:16:42.159
That is a mathematical summary, not a dependable forecast.
00:16:42.399 --> 00:16:50.399
Early stopping can exaggerate effects, surgical technique varies, and the trial does not specify important flow control details.
00:16:50.639 --> 00:16:58.320
A larger study may still find that the direction of risk is higher with the depot while estimating a substantially smaller difference.
00:16:58.559 --> 00:17:03.519
The risk cannot yet be estimated reliably across centers or devices.
00:17:03.840 --> 00:17:06.880
The most important practical implication is narrow.
00:17:07.119 --> 00:17:18.400
The findings argue against routine use of the tested 20 mg depot with Susanna UF surgery for refractory primary open-angle glaucoma until better safety evidence is available.
00:17:18.640 --> 00:17:30.720
They do not establish that 10 mg triumsinolone with an omid valve, 4 mg triumcinolone with trabeculectomy, or ordinary postoperative steroid drops carry the same risk.
00:17:31.039 --> 00:17:33.920
Can glaucoma surgery lower eye pressure too much?
00:17:34.160 --> 00:17:34.559
Yes.
00:17:34.720 --> 00:17:41.039
Glaucoma surgery aims to reduce pressure enough to protect the optic nerve, but lower is not automatically better.
00:17:41.200 --> 00:17:45.759
Very low pressure can cause the eye's internal structures to shift or fold.
00:17:45.920 --> 00:17:53.039
Complications can include choroidal detachment, a shallow front chamber, swelling or folds at the macula, and reduced vision.
00:17:53.200 --> 00:17:58.079
In some cases, pressure must be raised by changing or restricting the drainage pathway.
00:17:58.319 --> 00:18:01.599
American Academy of Ophthalmology review of hypotony.
00:18:02.079 --> 00:18:05.759
The Susanna trial changes our understanding in a focused way.
00:18:06.000 --> 00:18:19.039
It provides randomized evidence that a long-acting 20 mg triumpsinolone injection added to this non-valve device may bring a substantial late hypotony risk without improving average pressure control.
00:18:19.359 --> 00:18:28.960
It does not show that every steroid injection is unsafe, prove exactly how the effect occurred, or justify stopping routine postoperative steroid drops.
00:18:29.200 --> 00:18:36.559
It makes device design, tube flow, dose, and the ability to adjust treatment important questions for future trials.
00:18:36.799 --> 00:18:40.720
All links to sources are available in the text version of this article.
00:18:40.880 --> 00:18:44.480
You can find the full article at Visual Fieldtest.com.
00:18:44.640 --> 00:18:45.759
Thanks for listening.
00:18:46.000 --> 00:18:51.839
To check your visual field, click the link at the bottom of this article or visit visualfieldtest.com.