Wagner: Our topic today is retinopathy of prematurity (ROP): injection or laser treatment. This is a typical case of a 23-week-old premature infant who developed stage 3 ROP at the margin of zones I and II with plus disease and a few small scattered hemorrhages at the demarcation ridge. He is now 34 weeks' gestational age and he is still intubated and requires high oxygen therapy for pulmonary issues. What would be your thought process in deciding if you would treat right now and what factors would you consider?
Bhatt: Based on that history and the age, I would lean toward intravitreal anti-vascular endothelial growth factor (VEGF) treatments over laser in this case. The reason for favoring intravitreal treatment would be the really posterior disease. It's still at the zone I/zone II border, so I think that's a lot of retina to laser. Giving the patient a chance to vascularize further would be beneficial and decrease myopia in the future. The Bevacizumab Eliminates the Angiogenic Threat for Retinopathy of Prematurity (BEAT-ROP) study and some of the more recent studies have given us good data that bevacizumab is at least as good as, if not superior to, laser in these really posterior eyes.
Wagner: For general anesthesia, do you take the patient into the operating room or do you do it in the neonatal intensive care unit (NICU)?
Bhatt: The patient is already intubated, but at my hospital we would do general anesthesia for laser treatment. Our hospital requires the laser treatment to be done in the operating room due to safety concerns and they want a very controlled environment. But I would prefer to treat in the NICU.
Wagner: For an injection, do you have to go to the operating room?
Bhatt: We can administer an injection in the NICU. I use a hospital gown and a mask and sterile gloves, and I try to use a sterile set of instruments for each eye and prep with povidone-iodine. We don't use any sedation, just topical anesthetic on the eye. The patients tolerate it well and we've had good success with no infections so far.
Jordan: I would definitely do intravitreal treatment for this patient. We use bevacizumab 0.125 mg in a 0.1-mL syringe at our institution and that's worked well for us so far. We're waiting for results on the next phase of the Pediatric Eye Disease Investigator Group (PEDIG) study to consider a lower dose. We would do this in the NICU and the neonatologists usually prefer some sedation to make the patients more comfortable, such as midazolam and fentanyl. I also use povidone-iodine before and after the injection and a semi-sterile set-up for each eye, using a totally different set of instruments. Our institution also requires double gloves because it's technically a chemotherapy agent.For laser treatment, we also have to do it in the operating room due to safety issues. In this case, I agree that this disease is too posterior to warrant laser treatment at this time. But these patients need to be monitored closely and generally we do have to do laser treatment at some point, depending on whether or not they fully vascularize, if they're being discharged and there's risk of follow-up concerns, or if there's recurrence of the ROP.
Bhatt: At what age would you do laser treatment?
Jordan: We usually try to wait until 60 weeks, especially if they're staying in the NICU for a long time. They usually end up having another surgery (eg, tracheotomy, gastrointestinal tube, or hernia repair), so we try to coordinate the surgery scheduling to reduce their anesthesia time.
Wagner: As a matter of technique, do you grasp the eye with forceps when you give the injection? What do you use as your landmark as far as where to inject?
Jordan: I grasp the eye right at the limbus with a Colibri forceps and then I mark posterior to the limbus, approximately 1.5 mm. I grasp the eye, put a drop of povidone-iodine right on that spot that was marked with the caliper, and then do the injection. I have an assistant put another drop of povidone-iodine right after that and rub over the conjunctiva to seal it.
Bhatt: I also grasp the eye with a forceps. Interestingly, the distance from the limbus is variable around the country. The BEAT-ROP study used 2.5 mm, so that's what I did early in my practice. When the ROP I study came out, they were recommending anywhere between 1 and 2 mm, so I chose 1.5 mm. My pediatric retina colleague routinely makes sclerotomies at 0.5 mm from the limbus and has had no problems with the lens. I don't think there is a consensus yet, but currently I do 1.5 mm from the limbus.
Wagner: In a bilateral case, do you tend to inject temporally in the right and left eyes or temporally in one and nasally in the other?
Bhatt: Whatever is easiest. I usually end up somewhere inferior, between the 5- and 7-o'clock positions. I think grasping inferior and supraduction in the eye gives me better control and makes it even in both eyes, but I think you're probably fine anywhere you put it.
Jordan: I agree, wherever it is easiest in these small eyes. For me it's generally temporally, but sometimes I go infranasally in the left eye just because of positioning and where the continuous positive airway pressure machine is located.
Wagner: At my institution, injections are usually done by the retina surgeons and they have developed a preference for using ranibizumab because there's less systemic absorption and the half life is not as long. Have either of you used anything except bevacizumab?
Jordan: I have not used anything else.
Bhatt: No, I have not. Looking at the RAnibizumab Compared With Laser Therapy for the Treatment of INfants BOrn Prematurely With Retinopathy of Prematurity (RAINBOW) study data, I think their reactivation rate was higher than I would want with an anti-VEGF agent. That shorter half life may be good for systemic absorption, but it's not necessarily good for ROP because infants are still in a dangerous spot when the drug wears off and then it reactivates and you need additional treatment anyway. I think having a longer half life is actually an advantage of bevacizumab. As far as the systemic absorption and the developmental impact, I think that question remains to be answered.
Wagner: I mention it because in the cases that I've seen where ranibizumab was used, it seems like there's a critical time at 8 weeks after the initial injection where recurrence can occur to the point that you need to make the decision to inject again or to do a laser treatment, and most of the time they end up doing a laser treatment in those cases, too.One of the disadvantages with the laser is the follow-up time. I just started thinking about this because the 2019 novel coronavirus (COVID-19) pandemic has shown us that more frequent follow-up visits are not necessarily going to be easy to do in that critical 6 to 8 weeks. Would that come into play at all with you with making a decision as far as the treatment?
Bhatt: At this timepoint, no. I'm still routinely seeing my ROP patients in clinic. That's one of the few diagnoses that we're allowing to come in for an urgent visit, at least for the more established patients. It would definitely be something to consider, particularly if it's approximately 40 weeks or mid zone II. If I'm not sure that the bevacizumab is superior to laser, I might lean toward laser treatment now. But the first decision that needs to be made is whether there's a benefit to the eye itself by using bevacizumab, and then you deal with all of the other things afterwards.
Jordan: Maybe we do need to rethink our time to laser treatment for protection for the patients with persistent avascular retina. It might be good to get it done and not require so many visits for those in-patients, as well as the outpatients. Most of our patients who have injection seem to stay the NICU for a long time. It's definitely an interesting point that we probably will need to reevaluate so that we can limit some of the examinations that they go through.
Bhatt: That could also be important because we may not have the manpower to do these examinations. There are so few ROP screeners and it only takes a handful being quarantined or getting sick to have a major impact on the nation's ability to do this work. I think the more efficient we can make the work and the follow-up and reduce the number of visits the better at this point because there are so many questions as to what's going to happen in the next few months.
Wagner: As far as the management of these patients in general, do either of you have access to retinal imaging or even intravenous fluorescein angiography?
Jordan: I generally do intravenous fluorescein angiography before laser treatment because it's really helpful to see those areas of the persistent avascular retina. It also shows if there is any hidden neovascularization. We didn't always do it, but we have been doing it routinely more recently. We have to take the patients to the operating room anyway for the laser treatment, so it's convenient to be able to have the intravenous access. They're asleep, so it's not hard for them to have the device on their eye for a long time.
Bhatt: I do have access to it, but I don't use it that much. I know a lot of people use it and so I'm probably an outlier in that regard. I agree that it delineates the vascular-avascular junction well. I just don't think it changes my management. I still laser the same amount of avascular retina that I would have without it. I've never found any areas of capillary dropout, nonperfusion, or ischemia that make me want to do more posterior laser treatment than I was planning. Although it may show some areas of neovascularization or leakage, again, I don't think it matters because you're going to be lasering in that same encounter and it's going to help those things. I think it's probably most useful in infants in whom you don't plan to laser because you want to know what's happening at the far periphery and it bears some kind of sub-clinical activity still happening.
Wagner: I think we expect that there will be recurrences in some of these cases that end up getting laser treatment. How often do you think reactivation after anti-VEGF injection that may require additional treatment occurs?
Bhatt: I've looked at our data using the 0.25 mg dose and our need for re-treatment has been low (approximately 8%). I haven't had that much success with re-injection, so my second treatment is almost always laser. I think 6 of 6 eyes that I re-injected all needed laser in the end. Of the patients who reactivated, all of them did so by 45 weeks' postmenstrual age or earlier. I have had a few persistent avascular retina cases, but I think that's a separate issue. That's not really treatment per se, it's more prophylactic, but that's an even smaller percentage (2% or 3%).
Jordan: We actually just went through all of our data to present at the American Association for Pediatric Ophthalmology & Strabismus annual meeting. My fellow is the one presenting, so I don't know the numbers, but my impression is that if it's going to reactivate, it will be by 45 weeks. We recently had a patient who had an injection and he was doing fine for a long time, then he reactivated a little and had a successful laser treatment. Then 6 weeks later, he had a retinal detachment and had a vitrectomy and more laser treatment by our retinal surgeons and unfortunately still detached. It was one of the stranger cases we have seen here. So I think we just have to keep following these patients if they have laser treatment after bevacizumab injection. This is the one time I will not let them go long after because if the disease was really bad and really posterior originally, even laser treatment may not protect.
Wagner: I think that's an important point that even when we're doing the correct treatment and at the right time, they don't always respond the way we think that they should. That's why we always double counsel about the way we treat.
This Eye to Eye session was conducted on March 30, 2020.