Treating the Patient

Treating the Patient

Newer diagnostics are helping tailor more specific treatments for people with retinal disorders.

Highlights from a roundtable held during the Retina Society Meeting in Washington, D.C. on October 5, 2012. View video from the roundtable at

Peter Kaiser, MD: When a patient initially presents with AMD, what diagnostic tests are performed to verify the diagnosis?

Jeffrey Heier, MD: After obtaining the history and performing the exam, if we have a high suspicion for AMD, we will obtain fluorescein angiography (FA) and spectral domain OCT in the form of volume scans. The scans usually confirm neovascular AMD or bring to light variants such as vitelliform disease or high-risk drusen that are masquerading as AMD. I want a view of the entire macula to fully assess what is going on. It is too easy to miss disease with single-line scans.

David Brown, MD: Particularly in AMD, there is too much risk of selection bias if you only evaluate a few scans. If you do not look at all the best high-resolution scans, you are not giving your patient the best care he can receive.

Dr. Kaiser: What role does FA play in the current management of AMD and in determining lesion size and composition?

Karl Csaky, MD: If someone presents with visual complaints and I am not certain there is absence of fluid on the OCT scans, I will order FA. Even with volume scanning on the OCT, there is still a chance you could miss some fluid. The fluorescein can add information. For example, you can then see an area of increased hyperfluorescence with some leakage, which would indicate the presence of a neovascular complex. For me, FA still has an important role in making the diagnosis.

Dr. Heier: For new patients in my clinic, I order both OCT and FA in order to appropriately evaluate disease status. Our established patients already have SD-OCT by the time I see them, and if changes warrant further evaluation, we order the FA at that time.

Management Strategies

Dr. Csaky: I do not think the lesion composition is as important as it was in the past because the anti-VEGF injections seemed to work across all classifications. The only caveat to that is when I am led to believe there is polypoidal activity — where evidence suggests the anti-VEGF treatments are not as effective as photodynamic therapy (PDT). Many patients already know about the anti-VEGF medications and how effective they can be in some cases; I want to ensure that patients also know that we might make a switch to PDT earlier in their treatment regimen.

David Boyer, MD: For patients who have definite polypoidal disease, the EVEREST trial showed anti-VEGF therapy was visually equivalent to PDT but that the polyps seemed to close down better with PDT. In our hands, aflibercept (Eylea, Regeneron) for polypoidal disease may be better, but I am not yet certain why that is.

Dr. Heier: Also, in patients with a suspicion of vitelliform disease, I obtain autofluorescence to help show the lesion. It is a little more difficult to diagnose with fluorescein alone.

Dr. Kaiser: What imaging device(s) do you use for follow-up?

Dr. Boyer: I usually obtain an OCT and compare it to the previous scans. I typically utilize a treat-and-extend protocol, so I treat until dry and then I will extend it by 1 or 2 weeks depending on the status of the other eye. If after several injections the patient is not dry, I double the dose and bring him back in 2 weeks to see if I have missed a masquerade syndrome or if the lesion is non-VEGF responsive. I then run an indocyanine green angiography (ICG), enhanced depth OCT and autofluorescence.

What’s Important to Notice?

Dr. Kaiser: How do you define “wet” vs. “dry” and what do you look for on OCT to make that determination?

Dr. Boyer: To me, dry means there is no intraretinal cysts or subretinal fluid left as far as I can tell on OCT. I do not tolerate intraretinal fluid, and I certainly will not tolerate subretinal fluid. I may over-treat as a result, but if I can eliminate all the fluid, I go to an extend mode.

Dr. Kaiser: Where do you place subretinal pigment epithelial (sub-RPE) fluid within that classification scheme?

Dr. Boyer: I attempt to dry all eyes and in some cases, the sub-RPE fluid does go away. But if it persists after three or four injections, I move to a maintenance phase. Recently, I have been switching some of those non-responders to aflibercept because it seems it may close the detachments a little better.

Dr. Brown: New sub-RPE fluid, defined as darkness below the RP level that was not there on the previous scan, is always disease activity. Some large serous pigment epithelial detachments (PEDs) may or may not be related to wet AMD — they could just be a manifestation of their dry pump.

“We are probably underutilizing FA, based on CATT and HARBOR PRN data.”

David Brown, MD

Dr. Csaky: There is no set guideline that is applicable to a majority of patients with wet AMD. When is persistent fluid bad? I may tolerate a little subretinal fluid that remains despite several reinjections more than intraretinal fluid, which has been shown in the CATT study to be associated with more severe vision loss. It is those variations that necessitate individualized decisions. I am finding there is no ‘one size fits all’ treatment.

Dr. Boyer: The other eye’s status and the overall health of a patient need to be considered as well. Some patients just cannot come in every 4 weeks. Either they do not have the caretakers or some other reason prevents them from coming. If the other eye was already lost to hemorrhage, I would continue to treat the patient's other eye monthly.

Dr. Heier: CATT and VIEW tell us these patients are not drying, that the majority will have some degree of fluid that can range from a few intraretinal cysts to much more. For me, if there is a small amount of fluid, volume scans are essentially unchanged over several visits and they have not undergone a drug switching protocol, I may leave them where they are. But, if a patient has a PED and it gets smaller after each subsequent injection, I continue to treat to achieve an effect. However, if I treat and the fluid dissipates but the PED is stable, I will not go after it.

Dr. Kaiser: Does anyone use other imaging besides OCT during routine follow-up visits?

Dr. Heier: You need to compare to the previous scans utilizing volume scans to ensure you are not missing subtle disease.

Dr. Csaky: If patients are not doing as well as I would like, I might run another FA to see what is occuring with the neovascularization on a year-to-year comparison. There was a small subgroup of patients in ANCHOR and HARBOR where the neovascularization grew but visions remained stable or improved, even on monthly ranibizumab (Lucentis, Genentech).

Dr. Brown: We are probably underutilizing FA, based on CATT and HARBOR PRN data. In the latter, the lesions continued to grow, which you would not know without FA. I think of PRN as Progressive Retinal Neglect. It may be that even though we think we are doing the right thing with treat-and-extend, we may be doing ‘treat and extinguish.’ If we allow lesions to leak at all, they may grow over time. The studies alluded to a dry OCT but leakage still presents on FA. In my opinion, if extending treatment out to 8 or 10 weeks is being considered, an angiogram should be done before going any longer.

Dr. Heier: What if you do not see leakage?

Pharmacy Compounding
Issues surrounding accreditation and methodology remain, as do ideas on how to store the medications.
Dr. Heier: When issues about contamination in bevacizumab became public, the American Academy of Ophthalmology (AAO) and the American Society of Retina Specialists (ASRS) issued recommendations for pharmacy accreditation and methods that should be used, and some pharmacies did not meet those standards.
Dr. Csaky: In the CATT study, we were very involved in testing the quality of bevacizumab. We performed extensive testing on both bioactivity and aggregation. The FDA required a great deal of testing to be done, which some compounding pharmacies do not require.
Dr. Kaiser: Fractionation is the big safety issue with bevaciuzmab in my mind. You have lost control in terms of what you receive and deliver into the eye. Are there steps we can take to ensure the sterility, to ensure the drug did not undergo a freeze/thaw cycle, and to ensure no contaminants are present?
Dr. Heier: We have to ensure our compounding pharmacy follows the recommendations of AAO and ASRS.
Dr. Kaiser: What about storage? Any pearls?
Dr. Csaky: We keep everything from the compounding pharmacy at 4 degrees and store it for a maximum of 3 months. You can start losing some activity in a syringe with bevacizumab, so practices should be careful with how much they purchase.
Dr. Kaiser: At the Cole Eye Institute, our fractionation is done by our own pharmacy under a sterile hood and using all the standard pharmacy procedures. We send out random samples of syringes from every lot for microbiologic testing before we actually use that particular lot. Does anyone test their compounded bevacizumab?
Dr. Brown: Initially, we followed the Bascom Palmer pharmacy guidelines and sent some for culture, some for endotoxins.
Dr. Csaky: We use a pharmacy in San Francisco that follows those guidelines.

Dr. Brown: I find fluorescein very difficult to read after anti-VEGF therapy, whether there is leakage or staining. OCT has become an invaluable tool. However, late leakage on FA in the absence of OCT activity usually means that the lesions are leaking but the RPE pump is doing an acceptable job. If there is no activity visible on OCT and no late leakage on FA, one can consider watching observantly.

Different Dosing Regimens

Dr. Kaiser: What are your dosing regimens?

Dr. Csaky: If I have a one-eyed patient with a hemorrhage in the fellow eye, I tend to be aggressive and treat on-label or go out to 6 weeks at most between injections. For the routine patient, I will try to go beyond monthly injections or try to go beyond every 6 weeks.

Dr. Heier: About a third of my patients have disease in the fellow eye with poor vision, but not necessarily submacular hemorrhage. For those patients, I treat somewhere around a 4-6 week interval. It might occur that each time I extend to 6 weeks, they are brittle and leaking but at 5 weeks they are okay.

Treatment Decision-making

Dr. Kaiser: How do you decide what treatment to use in a treatment-naïve patient who has confirmed wet AMD?

Dr. Brown: I show the patient a healthy eye on OCT and then I show him his scan, explaining that the fluid and thickness is why he cannot see. I explain that he is going to need shots in his eye, and that we tend to use ranibizumab or aflibercept (I prefer FDA labeled treatment for patients, if possible), but insurance issues may necessitate using bevacizumab instead. We tell patients they will need at least three shots, and that 70-80% of patients will need injections for the rest of their lives.

Dr. Kaiser: Do you start your patients with FDA approved drugs? What about bevacizumab?

Dr. Heier: I start the majority of my patients on bevacizumab. I explain it is solely a cost issue, that it is an off-label use, but that the CATT study has shown both drugs do a good job. However, we have a very low threshold for switching them. If cost was not an issue, I would rarely use bevacizumab. In Massachusetts, we have received letters from insurers telling us that they have reviewed the results of CATT, that their experts have determined that the drugs are equivalent and they are asking us to use bevacizumab at least as first-line therapy.

“We tell patients they will need at least three shots, and that 70-80% of patients will need injections for the rest of their lives.”

David Brown, MD

Dr. Csaky: I believe a dry retina is a happy retina and if I can get that more effectively with aflibercept or ranibizumab — which the data support — I will use one of those first. If there is a co-pay insurance issue, then I will use bevacizumab.

Dr. Heier: The reality is that bevacizumab is effective for many patients. If a patient has the typical strong anti-VEGF response, then I have a very low threshold for moving them off. I probably start 80% of patients on bevacizumab, but my overall numbers are probably 50% because I move them off so quickly.

Dr. Brown: If it was not for the compounding pharmacy issues, I would agree. But you have to explain to patients that we have two approved drugs and one that is not approved, but is much cheaper. If insurance will pay for the approved drug, I feel more comfortable with that from a safety standpoint. From an efficacy standpoint, I absolutely agree with you — 50-70% of patients do great with monthly bevacizumab. But until we can have it in a form that is safe from endophthalmitis, I think you are exposing patients to an unnecessary risk if they have good insurance.

Dr. Boyer: I tend to start with bevacizumab, but again have a very low threshold because I utilize treat-and-extend and I want the drug that dries the eye most effectively, most quickly and for the longest duration of time. I have had patients extended out to 2-2.5 months with good vision and I have never moved them to another treatment. But if patients do not dry up after a couple of treatments and I do not see a positive response, then I immediately switch them.

Dr. Csaky: Overall, most of the neovascular AMD clinical trial data supports the notion that aflibercept and ranibizumab produce better anatomic results than bevacizumab, so I am a little less likely to use bevacizumab. My approach is to start out being as aggressive with getting whatever regression of the CNV and removal of fluid I can, so down the road they may need fewer injections. I am willing to try anything to get them on a schedule that will result in fewer injections later on.

Dr. Heier: The VIEW data suggests that aflibercept may actually dry the retina more thoroughly than ranibizumab. We have all had a large batch of suboptimal responders over the past 6 or 7 years. We had roughly 150 patients we switched to aflibercept and after a single injection, 75% had a significant anatomic response, but the visual response was small. Now, in a year we may have suboptimal aflibercept responders we will switch to ranibizumab, but the VIEW data suggests aflibercept may provide an advantage.

Dr. Kaiser: I currently use aflibercept in most of my treatment-naïve patients because I think it is identical to ranibizumab in terms of efficacy, and better in terms of cost and duration of action. In patients without insurance coverage or with financial/insurance issues, I start with bevacizumab.

Dr. Csaky: At least having aflibercept in the background allows you to offer something to the patient down the road. Dr. Heier‘s approach of having bevacizumab as a first-line, then ranibizumab, and then aflibercept as a step-wise approach needs to be considered. There is no data to suggest that any one approach is better than another.

Dr. Boyer: A study by Saada and colleagues showed bevacizumab failures who were switched to ranibizumab had an improvement,1 but the reverse was also true. We are going to have patients who are non-responders to aflibercept and may have to go back to ranibizumab, which may work.

Dr. Heier: We may all have patients who do better with bevacizumab than ranibizumab, but more often, it is the other way around, and there may be the same gradient with aflibercept and ranibizumab. The one thing we have overlooked is we have labeled usage saying we can treat every other month with aflibercept and have good outcomes.

Dr. Csaky: That is what has surprised me the most — the number of patients who need injections every 6 weeks, or even every 5 weeks instead of every 2 months. We have seen very inconsistent results with that 2-month window in our practice.

Dr. Brown: There may have been some selection bias because they were typically monthly ranibizumab patients we switched over. But even in the treatment-naïve patient, I think it is 20-30% who have to be treated more frequently.

Dr. Boyer: I find that many patients cannot extend to 2 months before experiencing leakage.

Dr. Heier: In suboptimal responders, I may treat with a double dose monthly of ranibizumab, but it is reasonable to expect they would require more therapy. We know from the VIEW results that only about 50% were dry. At 1 year, it stands to reason patients were fluctuating throughout even with monthly therapy. About 73% were dry at 1 year versus 60% on ranibizumab, so there is still a large number of patients who have fluid even with monthly therapy.

Dr. Brown: Ranibizumab has the most years of patient use. It has the shortest systemic half-life and the least anti-VEGF suppression. In diabetic patients (with DME or AMD), decreased VEGF receptors in the myocardium leads to silent ischemia and any little bit of VEGF suppression may potentially block the development of collaterals. If you worry about any of that, then you should start with ranibizumab.

“I currently use aflibercept in most of my treatment-naïve patients because I think it is identical to ranibizumab in terms of efficacy, and better in terms of cost and duration of action.”

Peter Kaiser, MD

Dr. Csaky: For patients who have been on ranibizumab for 5 to 6 years or longer, there has been no long-term retinal degeneration or long-term ill effects. At this time, we do not have that information with aflibercept and it has a much higher affinity for VEGF. It gives me a little room for pause, especially if someone has an existing geographic atrophy (GA) and presents with choroidal neovascularization. I try to avoid aflibercept in those patients. Again, we do not have the data, but it does give me pause.

Non-responders and Suboptimal Responders

Dr. Kaiser: How do you define a suboptimal responder?

Dr. Heier: Someone who has intraretinal or subretinal fluid, and that fluid has persisted following at least three monthly injections. That does not mean that they do not have good vision. There may be patients with 20/30 vision but persistent fluid. If I can get rid of the fluid, maybe there is a line or two of vision still to gain.

Dr. Brown: A non-responder is often a misdiagnosed central serous retinopathy, but maybe because the patient has drusen, we call it AMD. But a half-dose of PDT later and the retinal fluid is gone, and the patient does not need the injections.

Dr. Kaiser: How do you differentiate a non-responder from a suboptimal responder from a responder?

Dr. Boyer: After three injections without a response, I may double the dose (depending on what I am using) and have the patient return in 2 weeks. If there is absolutely no response, then I probably have the wrong diagnosis. If I do not see any VEGF response, I might switch to aflibercept or ranibizumab. If there is still no response, I start enhanced depth imaging. I will start autofluorescence, but I am looking for another diagnosis at that point.

Dr. Csaky: Some patients have extensive polypoidal disease, which is extremely recalcitrant to the anti-VEGF drugs. That is not necessarily a misdiagnosis, but it may require redirecting your therapy toward something else.

Dr. Heier: I learned years ago that ICG is clearly still useful, and some of the newer imaging devices can aid us in appreciating more of these masquerade syndromes like polypoidal or retinal angiomatous proliferation. Now I look at ICG very early on.

When Nothing Else Can Be Done

Dr. Kaiser: Is there ever a point where you will not treat because the eye is too far gone?

Dr. Csaky: In a disciform scar, I look for any degree of subretinal fluid or intraretinal fluid. You might not get them back to 20/40, but you can increase some of their ambulatory vision and may get to 20/80 even if they have significant fluid. If there is a dry disciform scar, I may try a single injection but I am much less optimistic. If there is fluid over a scar, some patients may benefit from the anti-VEGF medications.

Dr. Brown: I always try that, and it almost never works and patients cannot tell a difference between the two eyes.

Dr. Csaky: But patients always want to feel like everything was done.

Dr. Heier: I will give an injection and ask them for feedback. If patients cannot recognize a change, I see little point in continuing. There are some who notice or report a benefit, albeit a small benefit. Whether it is real or not is difficult to accurately ascertain, but even if it is subjective, I believe the emotional benefit can be helpful for patients who are already prone to depression.

Dr. Boyer: I think it is very hard to stop treatment knowing that if the patient comes back with a hemorrhage, he usually cannot regain his vision. A major hemorrhage is the kiss of death with this disease, so I tend to follow these patients forever. But we have to look at the whole patient and how they are functioning and try to avoid that major hemorrhage if their getaround eye is 20/400.

Dr. Csaky: You have to consider the visual response as well. If the retina is dry but they are at 20/200 you are going to be in a holding pattern to keep things stable. There will be a good number of patients who have an anatomic response but no visual improvement. ✦


1. Gasperini JL, Fawzi AA, Khondkaryan A, Lam L, Chong LP, Eliott D, Walsh AC, Hwang J, Sadda SR. Bevacizumab and ranibizumab tachyphylaxis in the treatment of choroidal neovascularisation. Br J Ophthalmol. 2012 Jan;96:14-20. Epub 2011 Jul 26.