Issue: July 25, 2018
July 20, 2018
14 min read

Round table: Consensus remains elusive how and when to treat ROP with anti-VEGF vs. laser

Issue: July 25, 2018
You've successfully added to your alerts. You will receive an email when new content is published.

Click Here to Manage Email Alerts

We were unable to process your request. Please try again later. If you continue to have this issue please contact

There are more questions than answers when pediatric ophthalmologists are faced with treating retinopathy of prematurity. A consensus is yet to be determined on when and whether to use anti-VEGF injections vs. laser, and when and how to treat recurrence.

A study in Ophthalmology identified persistent abnormalities on fluorescein angiography at 4 years after anti-VEGF treatment for ROP. The study sparked debate among members of the OSN Pediatrics/Strabismus Board, who gathered at the American Association for Pediatric Ophthalmology and Strabismus meeting in Washington for a round table discussion of hot topics led by Section Editor Robert S. Gold, MD.

Held jointly with the International Strabismological Association, the AAPOS meeting highlighted issues in strabismus surgery as well, and the OSN round table participants did the same, identifying their preferred incision — fornix or limbal — and their preferred approach — plication or resection — for eye muscle surgery.

One common thread weaves through the fabric of all the panel’s discussion topics: Management decisions continue to be individualized to the patient and the operator.

OSN Pediatrics/Strabismus Section Editor Robert S. Gold, MD, leads a panel discussion on how emerging study results are influencing management decisions involving treatment of ROP as well as surgical approaches to strabismus correction.

Source: Robert S. Gold, MD

Robert S. Gold, MD: The first topic that we are going to discuss is the article by Domenico Lepore, MD, and colleagues in Ophthalmology that talks about the 4-year outcomes of intravitreal Avastin (bevacizumab, Genentech) vs. laser for treatment of ROP and some of the concerns and issues that are associated with it.

R.V. Paul Chan, MD: This is a follow-up study that looks at the fluorescein angiographic findings in children who were treated with bevacizumab in one eye and laser in the other eye. The study concludes that, yes, there are angiographic changes even at the 4-year post-treatment time point. One question, however, is whether or not these angiographic findings are clinically significant.

Roberto Warman, MD: One reason to discuss this paper is for practicality. The majority of people who are screening ROP are not going to have access to fluorescein, and those who do have access cannot do fluorescein in all cases. It is not practical. So, we now know that we are missing clinically, frequently some type of process that is still active. However, we also know that most of our patients do not progress to retinal detachment. So, how comfortable can we be continuing doing it without fluorescein angiogram? That is the question.


Chan: I think many children who receive anti-VEGF treatment for ROP should get fluorescein angiograms (FAs) to follow them, especially in the long term, because it is not so uncommon to see changes and areas of persistent avascular retina.

We recently published a case in JAMAOphthalmology in which we performed an intravitreal injection of bevacizumab for ROP, and about 4 or 5 years after the injection, there was still a significant area of avascular retina that could not be seen on clinical exam but could be seen on the FA. So, we are seeing these cases. Also, we are doing more FAs now — because we can — but we do not yet have a good normative database of what a normal fluorescein angiogram is going to look like in children as they develop. There are a lot of things we do not know. I do think that these angiographic abnormalities, seen in children who received anti-VEGF, are real, and we should perform FAs if we can. But not everybody is going to have access, and these imaging devices are not inexpensive.

Rudolph S. Wagner, MD: I agree that it would be great if we could have FA done in all patients who were treated, but this is not practical since access is limited. For many people, it is perhaps easier to treat these cases with an injection than with a laser, and you are going to see more and more cases treated with injections, creating a large cohort of children that you would like to be able to study with angiogram. I think that the natural history of those patients treated with laser vs. injection is not established, especially for the patients who were injected. We need more data, but presently this will be difficult to obtain.

Kenneth P. Cheng, MD: Let me preface by saying that I was involved in all the initial studies, CRYO-ROP and ETROP, but I got out of the ROP field of expertise a number of years ago and so have no personal experience with injecting these eyes. Even so, what Rudy said is important, that injections are easier to do than laser and people have gotten comfortable with injecting these eyes; you see the disease stop and you can breathe easier quicker. But we still have questions. We do not know what these study findings mean. This is an issue that has to be studied more in depth with a randomized controlled clinical trial. That is the only way to move forward without taking excessive risk to patients long term. Even the results of such a study are going to be difficult because we may not know whether these large areas of avascular retina are going to be an issue for patients 20 years hence, and if a lot of these eyes could have been saved and stabilized for good central visual acuity for a lifetime as we know from all the previous work with laser, then we are going into uncharted territory.

Roundtable Participants

  • Robert S. Gold, MD
  • Moderator

  • Robert S. Gold
  • R.V. Paul Chan
  • Kenneth P. Cheng
  • Scott E. Olitsky
  • Rudolph S. Wagner
  • Roberto Warman
  • M. Edward Wilson


Gold: I am going to put on my risk management hat as the chair of OMIC’s ROP task force. Avastin is an ongoing treatment and an ongoing question. Not only do we have the question now of the relationship with avascularity down the road, but also we continue to have the question of the relationship with neurodevelopment, and these questions are unanswered. So, we are going down that road with more people probably doing injections. Initially we saw reactivation with injections, and we still do. How far do you follow these children? My concern, as Roberto pointed out, is that the availability to do a fluorescein angiogram in these children is still extremely limited. In some settings, such as a university setting, it is not difficult, but in the usual community NICU or later in an outpatient setting, if we do not have a way to have a cooperative child do a fluorescein angiogram, it could be a slippery slope.

Warman: If we need to re-treat, should we always do laser and be over and done with it?

Chan: My personal feeling is that if a patient has already progressed to zone II and has a recurrence, then I would most likely consider re-treating with laser. But one major issue we have in this discussion is that there are multiple opinions about what the definition of “recurrence” is. We do not have a consensus on the definition of “recurrence,” and we lack specific guidelines on what to do if there is “recurrence.” Recurrence after intravitreal anti-VEGF for ROP is still something that needs further investigation. In my opinion, if there is neovascularization that recurred or if there is plus disease, I would probably re-treat with laser. That is my preference. And I generally will do it after performing a FA to help identify any areas of retinal ischemia that need to be treated. In many of these children it is often hard to see the avascular retina well until after the FA is done.

Fluorescein angiograms give us a significant amount of information, but similar to the issues we have with defining and treating recurrence, we do not have a consensus on how the information provided by the FA guides our management. What do we do if we see certain findings on FA? Most published papers on FA for ROP are descriptive, but how the FA truly guides our management for ROP is yet to be determined.


Wagner: I think Paul is right on with this. There is a lot of data out there that needs to be gathered. In the past couple of months, I saw one patient who was injected and then at 10 weeks after the injection had a recurrence of ROP. That one was lasered and did well. The vessels had extended farther into the periphery. In another case, laser was done after anti-VEGF injections were done twice. Once again, the vessels had progressed far enough to the periphery, and laser was successful. Confounding this whole issue is that some people are starting to use Lucentis (ranibizumab, Genentech) in place of Avastin, and that might have a different effect or a different time course.

Scott E. Olitsky, MD: Ultimately, the thing to remember is that we are fundamentally changing the way this disease behaves, and there are a lot more questions than answers at this point.

Chan: I agree. There are so many questions that still need to be answered, even after 12 years of publications on the use of these drugs for ROP.

Cheng: Paul, can you go through three scenarios: the ideal use of Avastin; the “I’m not sure if I should use Avastin”; and the “this patient should absolutely have laser instead”?

Chan: It is very user-dependent. I tend to consider using anti-VEGF in very aggressive cases such as aggressive posterior ROP or zone I disease that requires treatment. Even in zone I disease, if it is not involving the fovea, I may still prefer to laser it. Aggressive disease is when I have that conversation about anti-VEGF with the parents, and now I also discuss ranibizumab vs. bevacizumab.

For zone II disease that is not so aggressive looking but has plus disease in stage 3, I will tend to use laser as primary treatment. And, in cases in which there is a significant amount of iris neovascularization where I cannot get a good view to perform sufficient laser, I think that anti-VEGF is a reasonable option.

These drugs are here and available, and they are a great option in cases where we know laser will not do well for various reasons. Even though there are a lot of questions about the use of these drugs, they work well in promoting regression of ROP. But I think we have to be selective; like so many things, it is all about patient selection.


Gold: Paul said something important: It is a user-dependent determination. There are guidelines, but people do what they feel most comfortable doing.

Cheng: From my perspective, even after saving a lot of eyes, there were still eyes that you knew were not going to do well with laser — the bad zone I eyes and the iris neovascularization eyes. It is a wonderful thing that we have something we can do for those patients who before would not have done well.

Incisions for strabismus surgery

Gold: Next is the topic of strabismus surgery. We can start the conversation with your preferred incision for strabismus surgery.

M. Edward Wilson, MD: Any of the popular incisions for strabismus surgery are fine. Fortunately, the conjunctiva heals well, so surgeons usually use the incision they feel most comfortable with. I do virtually everything through a fornix incision, even reoperations. That is my preference. I would rather work through a fornix incision, even if a previous surgery was done with a limbal incision. I would prefer to dissect the conjunctiva from behind using a fornix incision. In elderly adults in whom the conjunctiva tears frequently, I have increased my usage of two radial openings in the conjunctiva beside the muscle, which was popularized by the minimally invasive strabismus surgery techniques, but I still do a fornix incision most often, even in these elderly patients.

Wagner: I do what I feel is best for the individual patient, and sometimes it is easier to do a limbal incision if you have limited visualization in a small child with a medial rectus recession. Sometimes the visualization is better if you can expose the area to a greater degree. As a good strabismus surgeon, you have to be able to use all of the incisions. I prefer a fornix incision for lateral rectus muscles because I like the way they heal. It is a matter of preference and what works for you and what you are trying to accomplish.

Cheng: I cannot remember the last time I did a limbal incision reoperation or virgin case. They are all from the cul-de-sac. And 95% of them, depending on whether I am working with a resident or not, do not need to be closed or do not need to be sutured, and they heal without any problems.

Warman: I was trained both in residency and fellowship exclusively with limbal incision. Then I joined partners who only did fornix incisions, and they taught me. I mostly have done fornix incisions since. It is rare that I would go back to doing a limbus. It would have to be a complex reoperation, a lost muscle type of procedure.


Olitsky: Mainly fornix, but limbal incisions for some of the reasons other people have said. I became less and less dependent on an assistant and do most surgery by myself. So in a case in which I might want a little bit more exposure, especially without having an assistant working with me, I would use a limbal incision. I agree that being able to use all of the options is important.

Wilson: Ken mentioned that fornix incisions usually do not need to be closed, and I agree with that, but I am in the habit of closing them all. I use a buried plain gut suture to avoid patients calling because they see the gap in the conjunctiva during healing. Even though we have not studied this, we do a lot of cases and my feeling has been that since we started closing all the incisions, we have gotten less pyogenic granulomas and other sorts of things that sometimes, rarely, come about in the area of the incision. So now I am in the habit of using plain gut suture and closing them all.

Wagner: I agree with Ed on this. I have seen cases, especially in what I call these peak Tenon capsule age groups, when a child is 3 or 4 years old, in which you can get some prolapse if the incision is not closed. This can lead to a granuloma formation. So, I like to close them, also.

Cheng: As I said, I do not close the overwhelming majority of mine, but if for some reason or other it looks like the Tenon’s might stick up, it is simple enough to throw in a suture. The one thing I do that may be a little different is that, if I do have to close it, I close it with a little bit of my leftover Vicryl from the case. There is never a complaint. It is underneath the lid, and it falls out. So, save that piece of gut or whatever you use.

Gold: We have talked in this forum about resection vs. plication before, but it is still a hot topic. Plication was becoming a little more popular, and then some individuals said they were not getting the results they wanted with plication, so they went back to resection. Personally, I have been comfortable with resections for a long time, and being in practice for as long as I have, I stay with something that I am comfortable with. I do not have any difficulty with resections, and I continue to do resections while some of my partners are doing plications and not having many issues. What are your comments and preferences?


Wilson: I still prefer plications over resections. I think when we discussed this previously, there were some anecdotal reports that seemed to indicate plications were not as strong as resections, either initially or later. I have not found that to be the case. With plications, you have to be careful to get the suture tight. When the muscle is being plicated, it is easy to have some conjunctiva or Tenon’s fascia prevent your two needle passes from coming together. In resections, we have all been used to tightening them. With plications, it is the same thing. It is a newer technique, and people have to make sure that they are getting it tight. In my experience so far, it is just as good, even long term, as resections and there is less surgical trauma. And you are not going to lose the muscle. It is easier for me to teach it to a resident — at least the way I do it — and so I still prefer it.

Wagner: I still do primarily resections for some of the same reasons that Bob mentioned. But if I am concerned about anterior segment ischemia, I would use plication that way. I think it is equally effective. I have seen some cases in which there has been an initial elevation postoperatively close to the limbus following a medial rectus plication. Usually the bump resolves as the muscle flattens over time. Because of familiarity, though, I continue to do mostly resections.

Cheng: Resections have worked well for me for many years, and I am comfortable with them. I jumped on the bandwagon; I played with plications. I have done a number of plications now, and I have not seen any change in the short-term follow-up of a couple of years’ results, but I do not find the speed or the ease of it to be an advantage to me. So, my preference now is still for resections. If the case looks like it is going to be something odd, in which verticals might be needed later or we need to save the ciliary vessels, then I will throw that case in the plication camp.

I am a little bit more reluctant to do medial rectus plications because of the thickness of the muscle and the bulk, but that does seem to disappear after a while.

A plication is also a good thing for a small effect, where I would otherwise worry that, because of the size of the clamp and whatnot, I could not achieve resection smaller than 3.5 mm.


Warman: I was comfortable with my resections, but after we discussed this topic previously, I decided to do plications and I did OK. But then came the recent paper by Alkaharashi and Hunter with 5-year data that are extremely skewed against plications now. I do not know that we have the answer yet as to whether they are equal or not. The plications have worked fine for me, but I do mostly resections because I have done them for a long time.

Olitsky: I remember that conversation and listening to Ed’s enthusiasm. In our group, there was a large spike in plications and I would say it is still fairly popular. A few people have switched back because of the long-term data, although I am not sure that is fair because we do not have an answer based on that. I also remember wondering, if plication is that good, are we going to start doing those in place of recession? I am curious, Ed, if plication is a procedure you would choose over recession. Are you doing those now instead?

Wilson: Yes, at least considering it. To recap, if you need to do two-muscle surgery and want the least surgical trauma, then a procedure in which there is no detachment of muscles — you are just folding them over — is the least surgical trauma, so I do consider that.

I work in a teaching environment, and I have found that residents and fellows can learn to do a plication that I would grade as excellent in fewer cases than learning to do a resection. Resection is a little harder. For those of you who do them so automatically, there may be no need to change, but the surgical trauma issue in a training environment is important to me.

Some of the issues in Hunter’s paper may be technique related. More of us have to publish comparative 5-year results before we will have answers.

Wagner: I think that one of the reasons that this has not been adopted universally or more frequently is that the techniques do vary. I do not think there is a standard technique for plication. When that is determined and taught, I think you will see a lot more use of it.


Disclosures: The round table participants report no relevant financial disclosures.