This roundtable discussion with Sophie J. Bakri, MD, Mathew W. MacCumber, MD, PhD, Angelo P. Tanna, MD, and Inder Paul Singh, MD, was moderated by Peter K. Kaiser, MD.
Dr. Kaiser: There have been several reports noting sustained increases in intraocular pressure after anti-VEGF injections. This panel, which consists of 3 retina specialist and 2 glaucoma specialists, will discuss this issue. Dr. Bakri, you were one of the first to describe this issue. Walk us back in time. What made you look at this? Was there a certain patient that really impressed you, that made you look into this idea that IOP can be increased after anti-VEGF injections?
Dr. Bakri: Back when we had been doing these injections for a couple of years, I started to notice on a couple of my patients that the pressure was high. Whenever the pressure is high, with no other reason you start to wonder why. So, I looked back. Do they have glaucoma? Have they been treated? Are they on drops? And I was finding nothing.
And so then I would do gonioscopy, etc., and then refer them to the glaucoma service. So, after I saw several patients with the same pressure issue after injections, I spoke to a couple of my colleagues and they had seen a couple of similar patients as well. The first patients were picked for a case report because they had no prior history of glaucoma, were not on steroids, and no family history of glaucoma. We found they had had nothing but anti-VEGF and the pressure was going up. And so that’s what started all this.
And others were noticing similar things and writing case reports. And then came the era of subanalyses; I looked at the MARINA and ANCHOR studies and found that although they looked at the mean pressure before and after treatment at 24 months, they didn’t look at whether there was a subset of patients that were prone to develop this. And as you know, you could potentially look at a study on a weak steroid and you could find that the mean pressure may go up just by a little bit, or maybe not at all. But on the other hand, there’s a subset of 30% of patients that may develop a significant increase in pressure. And we were more interested in whether there was a subgroup of patients that may be more prone to getting this after anti-VEGF.
Dr. Kaiser: Dr. MacCumber, how often do you see this? Is this something that you’re concerned about and do you actually describe it to your patients as a risk when you start anti-VEGF injections?
Dr. MacCumber: We’ve noted eye pressure rises, and we were aware of Dr. Bakri’s work and others’ as well. So we stepped back. We had access to the IRIS Registry, the largest ophthalmic database run by the American Academy of Ophthalmology, and we found enough data there to look at whether there have been sustained pressure rises in patients who have received anti-VEGF injections.
We looked at all the patients who received anti-VEGF injections for the main diagnosis AMD, vein occlusion, or diabetic macular edema, and filtering it down to those who had at least 1-year follow-up, and at least 1 injection, we got about 23,000 patients. Then we only looked at those who received only 1 drug, either aflibercept, bevacizumab, or ranibizumab. And what we found was that that the rates were not quite as high as those reported before. Now, this is real-world data so they’re not getting as many injections as patients received in clinical trials.
We found that, actually, pressure went down on average by about 0.5 mmHg overall. However, if the patients had a preexisting diagnosis of glaucoma, then they went up a little, on average about 1 mmHg. We next looked at the subgroup that Dr. Bakri described previously, who are patients who experienced at least a 6-mmHg pressure rise to more than 21 mmHg (clinically significant pressure rise). There we found some differences between the drugs. Overall for ranibizumab and aflibercept, there was a rate from 1% to 3% pressure rise.
However, with bevacizumab, regardless of the amount of injections, IOP went up in higher percentage of eyes than the other drugs. We examined at least 12 or more injections, 18 or more injections, and 25 or more injections of bevacizumab. The rate kept going up the more injections they received until for those who received 25 or more injections, the rate was up to 9.5% of patients that had clinically significant pressure rise, where, again, it was only about 1% to 3% for the other drugs. We only looked at eyes that had anti-VEGF injections. We did not look at a case control group. That could be another study we could follow up with.
For now, the only thing that we can really discern comparatively is the different drugs. So, I think it is something of concern, but not quite as much as we had feared from the data we had seen previously, particularly from the clinical trial data that Dr. Bakri described.
Dr. Tanna: Dr. MacCumber, that finding is really interesting, because it’s consistent with some basic science data that has demonstrated that bevacizumab, but not ranibizumab, is toxic to trabecular meshwork endothelial cells in culture, or at least reduces the metabolic activity of that cell population. So, there may be some direct biological effect of the anti-VEGF agent itself on the trabecular meshwork.1
Dr. Kaiser: Dr. MacCumber, over what time frame did these patients receive anti-VEGF therapy?
Dr. MacCumber: On average, it was about 1.8 years. The database only started in 2013, so we didn’t have many years of follow-up. We can only have up to 3 years to 4 years of data.
Dr. Bakri: I think one difference between the study that Dr. MacCumber did and other subanalyses of trials is that yours was real world, and I think that by 2013, most of us had started doing more treat-and-extend. The data from the trials are based on patients receiving monthly injections, and I don’t believe many of us do that on a routine basis for an extended period of time. So, that could be a reason the incidence has come down since then.
A recent survey showed that about half of retina specialists believe this is true2 and certainly this isn’t of the magnitude that we see with steroids. But there’s clearly a subgroup of patients who can be predisposed to getting an IOP rise. It’s unfortunate we don’t know how to predict that just yet, short of patients having prior glaucoma. So, I think the numbers are small, but I think there’s something there in a subgroup and hopefully, it will be up to the glaucoma specialists to figure out if this is a genetic predisposition or a toxicity to the meshwork.
Dr. Kaiser: In addition to the possibility of direct drug toxicity, there has also been a suggestion that a reason this may occur is higher injection volume, or a rapid injection technique where you have a very rapid rise in IOP leading to damage to the trabecular meshwork. Dr. Tanna, is this a plausible explanation, or is this a retina specialist’s explanation?
Dr. Tanna: There is no evidence that transient hydrostatic pressure elevation damages the trabecular meshwork. The study you refer to by Dr. Yannuzzi2 was a survey of retina specialists that showed that more frequent transient IOP spikes were associated with the use of a rapid injection technique and the use of higher volumes. Since it was a survey of techniques utilized by physicians and their observations regarding the frequency of IOP spikes, it is subject to recall bias.
Dr. Kaiser: What other risk factors should we be concerned about when giving anti-VEGF injections?
Dr. Singh: In my opinion, going along with what Dr. Bakri said, I’m definitely seeing this much less frequently than I used to several years ago, and I’ve attributed it to more frequent utilization of aflibercept and ranibizumab rather than bevacizumab. Does that fit with the retina specialists’ understanding of the utilization of these agents?
Dr. Kaiser: Let’s ask Dr. MacCumber that question, because he has access to the AAO IRIS Registry. Is the use of bevacizumab going down in the United States, or going up?
Dr. MacCumber: If you look at the patients when they start anti-VEGF therapy, use of bevacizumab is still quite high. But I think that over time, many of those patients are converted to aflibercept and less so to ranibizumab. So if you look at patients who have received quite a number of injections, I think it’s more common now to have a smaller percentage of those getting bevacizumab.
Dr. Tanna: Compounded bevacizumab obtained from some sources was found to have been contaminated with silicone oil microdroplets and protein aggregates. Silicone oil is used in the manufacturing process of some syringes. The study found freezing and thawing of the bevacizumab was associated with the presence of increased silicone oil microdroplets. These contaminants were thought to increase outflow resistance and increase intraocular pressure.3
Dr. Kaiser: From a glaucoma standpoint, how difficult is it to treat this IOP rise?
Dr. Tanna: The patients who are referred to glaucoma specialists are obviously more likely to require surgical intervention, so it is difficult to know based on clinical experience whether IOP elevation in these patients is more difficult to control. Retina specialists are not shy about starting ocular hypotensive therapy in these patients; so many of them are managed without incident and without being under the glaucoma specialist’s spotlight.
A population-based case control study conducted in British Columbia demonstrated that eyes that had 7 or more bevacizumab injections per year had a significantly increased risk of undergoing incisional glaucoma surgery. That’s telling, but only for bevacizumab.4
Dr. Singh: I tend to agree with Dr. Tanna. There is no doubt that the patients I see have already been on treatment for ocular hypertension. They’re more difficult from that perspective.
My local retina colleagues will start aggressive treatment with anti-ocular hypertensives early on when they see pressure rise, and those patients tend to be better controlled and need surgery less. I’ve seen a few consults where they’ve started earlier and I’ve been able to manage it medically and not have to go into surgery.
Dr. Tanna: So, Dr. Singh, do you think that patients don’t develop damage to their optic nerve, and they don’t get to an advanced stage where very low pressures are required, or do you think it’s because controlling the pressure early prevents further damage to the trabecular meshwork?
Dr. Singh: I think the higher the pressures for a longer period of time, the greater the potential for a collapse of Schlemm’s canal and as a result, decrease outflow and further collapse of the distal collector channels. So, in my opinion, when retina doctors start aggressive treatment and can control the IOP early on, those patients tend not to need as many medications to get their pressures down long term. Also, those patients tend not to develop as aggressive nerve damage and subsequently do not ending up needing as low target pressures. So, it’s kind of all of the above. I think that earlier treatment means reduced chance of disease progression. High pressure can damage the trabecular meshwork, so if you can control it earlier, you have less canal collapse and less distal collector channel issues.
Dr. Kaiser: That’s very interesting because as retina specialists, in general, we don’t pay too much attention to pressure. But it sounds like as it starts to creep up we need to be more vigilant and aggressive with our management. At what stage would you recommend a retina specialist start IOP-lowering drops, let’s say in a patient who has no family history of glaucoma and no risk factors for glaucoma?
Dr. Singh: Although there are many other risk factors to consider, I think if we start to see the IOP creep up consistently more than 4 mmHg to 5 mmHg, even if they have a healthy nerve, if there is asymmetry between two eyes, I would have no hesitation to start treatment.
The idea is to prevent damage from occurring well before there is functional loss and need for even more aggressive IOP reduction. The more nerve damage you have, the lower the IOP needs to be to protect the remaining nerve fibers. Usually, the earlier the treatment, the easier it is to halt the progression.
Dr. Kaiser: Is there a medication class that’s better to start with, or is it like any other glaucoma, start with the prostaglandins first?
Dr. Singh: The prostaglandins have a high safety and compliance profile and on average demonstrate a 30% reduction, so I usually do add a PGA first. Although I am usually not concerned as much about macular edema, if there is a high risk, I may choose an aqueous suppressant, which can be quite powerful immediately, especially in combination form. But I would say if it’s easier for a doctor, a prostaglandin is fine with me unless there’s some issue. Again, I’m not really worried about CME or any inflammatory side effects of the PGA. Dr. Tanna, any thoughts on that?
Dr. Tanna: I would be a little bit more conservative about the initiation of therapy. One thing that I think needs to be taken into account, which retina specialists are not used to looking at, is the central corneal thickness. In patients who have very thin corneas, we have an artifactually low IOP measurement, so the real pressure is higher than the measured pressure.
We believe patients with thin corneas likely have a lamina cribrosa structure that may render the optic nerve itself more susceptible to higher pressure, and the opposite is true for patients with thick corneas. So, it would be very hard to make a decision about a patient who develops ocular hypertension on anti-VEGF therapy regarding a recommendation to treat or not treat without knowledge about the central corneal thickness.
Dr. Kaiser: So, Dr. Bakri, do you worry about a patient who has baseline history of glaucoma and they’ve now developed macular degeneration or retinal vein occlusion, does that change in any way how you manage these patients given all that we’ve talked about?
Dr. Bakri: In a way, yes. If you look at all the classes of drugs, I still believe that anti-VEGFs are the safest in terms of IOP. There are some things that one has to be aware of. If I were to do a series of injections on a patient without seeing him every month and doing an OCT, I would check his vision and pressure. So, I always check pressure every time that I see the patient. So, in that way, it has changed my management.
Dr. Kaiser: Dr. MacCumber, we’ve talked about the possibility from the IRIS Registry that multiple injections of bevacizumab may have a slightly higher risk of IOP elevation. If you have a patient with glaucoma or risk factors for glaucoma, do you take that into account or start with bevacizumab anyhow?
Dr. MacCumber: I agree with Dr. Bakri. I think the pressure measurement is important every time, particularly for patients with preexisting glaucoma, and I think that we should consider topical glaucoma therapy if the pressures are reproducibly running high, taking into account the corneal thickness as Dr. Tanna had mentioned. If the patient has a preexisting history of glaucoma, I may be more apt to use aflibercept or ranibizumab than bevacizumab.
Dr. Tanna: I strongly agree with that approach.
Dr. Kaiser: Is there any role for checking pressure immediately after intravitreal injections? In clinical studies, we routinely check the pressures immediately after, and that’s why acute IOP rise is in every label because obviously, the pressure is going to go up when you give an intravitreal injection. I personally don’t check the IOP immediately after injection unless the patient loses vision. Dr. MacCumber or Dr. Bakri, do you still routinely check pressures after injections?
Dr. Bakri: We stopped that practice maybe a decade ago. But having said that, there is a rare patient, as you said, Dr. Kaiser, who would lose vision after the injection, and they’re typically glaucoma patients. And in those patients, I would check the pressure. But, I would also consider doing an anterior chamber tap before the injection. But that’s an extremely rare patient. Sometimes I may inject slightly less medication.
Dr. MacCumber: In the past, we were using steroid at .1 cc and these days, almost all of our anti-VEGFs are .05 cc and we’re not seeing as high a pressure rise. It’s unusual for patients to lose vision after injection. I usually check finger counting. Typically, they have counting fingers, if not immediately after the injection then within a minute or so. I treat similarly to Dr. Bakri.
I’d like to add that the research with the IRIS Registry is only going to increase and if anyone has a good idea regarding a retina or glaucoma problem, then present it to the AAO for consideration. I think it’s going to be a great resource in the future.
Dr. Singh: This has been a great collaborative effort between glaucoma and retina and very important to help raise awareness. So, great job to everybody.
Dr. Kaiser: We’ve gotten some great comments here, thank you to all my panelists.
- Kahook MY, Ammar DA. In vitro effects of antivascular endothelial growth factors on cultured human trabecular meshwork cells. J Glaucoma. 2010;19(7):437-441.
- Yannuzzi NA, Patel SN, Bhavsar KV, Sugiguchi F, Freund KB. Predictors of sustained intraocular pressure elevation in eyes receiving intravitreal anti-vascular endothelial growth factor therapy. Am J Ophthalmol. 2014;158(2):319-327.e2.
- Liu L, Ammar DA, Ross LA, Mandava N, Kahook MY, Carpenter JF. Silicone oil microdroplets and protein aggregates in repackaged bevacizumab and ranibizumab: effects of long-term storage and product mishandling. Invest Ophthalmol Vis Sci. 2011; 52(2):1023-1034.
- Eadie BD, Etminan M, Carleton BC, Maberley DA, Mikelberg FS. Association of repeated intravitreous bevacizumab injections with risk for glaucoma surgery. JAMA Ophthalmol. 2017;135(4):363-368.