Are Ranibizumab and Radiation a Winning Combination?

Studies show promising results for neovascular macular degeneration

Are Ranibizumab and Radiation a Winning Combination?

Studies show promising results for neovascular macular degeneration.

Robert Murphy, Contributing Editor

Irrefutable clinical and scientific evidence indicates that the single best treatment for neovascular macular degeneration is ranibizumab delivered monthly without ceasing. But we then come upon a trade-off between maximal efficacy and the treatment burden that attaches both to the patient's time and effort as well as the practice's schedule and administrative overhead.

Studies assessing the feasibility of reducing the frequency of ranibizumab injections have been largely disappointing. Genentech's HORIZON trial, for example, found that patients receiving injections monthly for two years saw their visual acuity plummet after switching to PRN treatment. Other studies have consistently shown three distinct groups of patients: those who respond and maintain, those who respond and deteriorate, and those who never respond at all. Altogether, about half of these patients would benefit from further intervention.

This issue is why clinical researchers are investigating the use of a combination regimen involving concomitant use of radiation therapy and anti-VEGF medication with a protocol that calls for fewer injections over the long term. These studies look promising. One approach is that of NeoVista, a Newark, CA, company that makes a surgical device designed for intraocular epimacular brachytherapy to deliver beta radiation directly to the targeted site. Across town, Oraya Therapeutics, has developed a device that delivers low-voltage X-ray radiation externally to the macula.

Ongoing studies show that both methods, when used with ranibizumab at baseline and PRN thereafter, allow for fewer subsequent injections over 12 to 18 months while yielding favorable and durable visual outcomes with a benevolent safety profile. Time will tell which of two methods works best. Preliminary data suggest that there may be a place for both. In any event, FDA approval of radiation therapy for wet AMD promises to relieve the treatment burden that now weighs upon both your patients and your practice.

In February, two clinical researchers right in the thick of things—Pravin Dugel, MD, of Retinal Consultants of Arizona in Phoenix, and Darius Moshfeghi, MD, of the Stanford University School of Medicine—reviewed their findings at the Retinal Physician Symposium in Las Vegas. Dr. Dugel is among the clinical investigators working with NeoVista to look at the safety and efficacy of epimacular brachytherapy used in combination with ranibizumab. Dr. Moshfeghi is involved in clinical trials examining Oraya's low-voltage X-ray radiation device used with ranibizumab. Here's a synopsis of their presentations.

Ranibizumab is highly effective for wet AMD, but chronic monthly injections are considered unsustainable long-term. Radiation could be an effective adjunct.


There's no way to change your practice mix. Likewise, there's no way to change radically the way you practice. When it comes to managing neovascular AMD, retina specialists are at the mercy of the injection code and the OCT code. Anytime government officials decide to change these components, everything else that you do likewise becomes vulnerable. And there's not a thing you can do about it.

There may be a way out of this bind. Studies show that the combination of anti-VEGF treatment with radiation is more effective than either therapy alone. Epimacular brachytherapy for neovascular AMD offers a beneficial synergistic triad of increased oxygenation from the vitrectomy procedure, targeted brachytherapy, and anti-VEGF treatment.

Many studies have addressed treatment-naïve patients. However, the patients who bear the brunt of what has become an unsustainable monotherapy model are those who have persistent fluid despite numerous anti-VEGF injections. The MERITAGE study in which Dr. Dugel has been involved was designed to address this population. Its goal was to alleviate the treatment burden by lowering the number of injections while maintaining visual acuity.

The clinical researchers wanted to ensure that the 53 patients received the full effect of anti-VEGF treatment. So they had two phases of ranibizumab treatment: a loading phase consisting of three monthly injections, and a maintenance phase of at least five further injections in the 12 months before enrollment or at least three additional injections in the six months prior to that time.

Patients had received up to 38 injections at baseline, with a mean of 12.3. However, although the number of injections was increasing prior to enrollment, their visual acuity was actually worsening. These were not easy patients. The retreatment criteria encompassed clinical, OCT and fluorescein angiographic parameters. The one used most commonly was a change in subfoveal thickness of 50 µm or greater on OCT.

The study's safety arm showed cataract progression in 86% of patients, presumably due solely to the vitrectomy. There were no cases of radiation-based retinopathy at one year. As for the main goal—to ease the treatment burden by reducing the number of injections—the mean number of injections dropped from 12.3 prior to epimacular brachytherapy to 3.9 (including the one required by the protocol at baseline) in the 12 subsequent months. One-quarter of patients remained injection-free at 12 months.

NeoVista's epimacular brachytherapy uses intraoperative radiation therapy to extend the interval between anti-VEGF injections.

As of early May, the safety and efficacy outcomes remained essentially unchanged. Of the 600+ patients worldwide who have received epimacular brachytherapy in combination with ranibizumab injections, only eight experienced vascular changes. Many of these patients had concurrent disease, including Lyme disease and malignant hypertension. Five of them subsequently resolved, signaling that their changes were not due to radiation. The vascular changes in the other three cases were mild and found to be nonproliferative. These patients have done well with no need for retreatment.

It looks like they achieved their goal of easing the treatment burden by reducing the number of injections. But they got an added pleasant surprise. Almost 50% of patients saw their vision improve—despite a difficult population who were losing vision with an increasing number of injections, and despite the fact that some of them had been treated for more than five years with up to 35 injections. Eighty percent of patients remained visually stable or improved. Those who improved enjoyed a mean gain of 7.5 letters.

A fair question: Was the visual improvement due to the anti-VEGF injections or the radiation? Interestingly, those who improved the most had the fewest injections. This fact suggests that the radiation must have had a significant beneficial effect. Patients with a classic component had a dramatic decrease in the size of the neovascular membrane. Two-thirds of these patients showed no evidence of the classic component by fluorescein angiography at one year. All those who had a predominantly classic lesion at baseline exhibited no evidence of the lesion.

How will epimacular brachytherapy be used? Its most common use will likely be for those receiving numerous ranibizumab injections, since they're the ones with the greatest treatment burden, as well as patients with pigment epithelial detachments, RAP lesions, and PCV-type lesions.

NeoVista has now launched the CABERNET study involving 492 patients—the largest device trial ever undertaken. Investigators will present their data in October at the AAO's Subspecialty Day Meeting.


As the folks at NeoVista were developing their beta-radiation device, most clinicians thought that the only accurate, precise and reproducible way to deliver radiation to the center of the macula was through a surgical approach. Oraya later came up with a highly collimated low-voltage X-ray unit that delivers its beams to the center of the macula externally in a nonsurgical approach.

The procedure calls for an I-Guide to be placed on the ocular surface with low suction to hold the eye in place. So-called fiducials—small silver tracking items—are kept on the I-Guide to allow for real-time tracking. The operator then administers three low-voltage X-ray beams through the inferior pars plana precisely to the macula while avoiding the optic nerve and lens. The operator's radiation exposure is negligible.

At the Retinal Physician Symposium, Dr. Moshfeghi reviewed phase 1 clinical results from Mexico City of the X-ray radiation device used in combination with ranibizumab. The study looked at the regimen's safety and preliminary efficacy and the results from different dosing strategies.

Inclusion criteria. Eligible patients were 50 or older with subfoveal choroidal neovascular membranes secondary to AMD. Visual acuity ranged from 24 to 69 letters on the ETDRS chart. Two cohorts included previously treated patients—the bulk of those whom we now treat—and those who were treatment-naïve.

Treatment protocol. The clinical researchers looked at three different treatment strategies. Two groups received a loading dose of ranibizumab at baseline followed by a second dose at one month, with radiation of either 16 or 24 grays (Gy) during the first 14 days; these first two groups received subsequent PRN ranibizumab injections monthly out to 24 months. A third group underwent 16 Gy at baseline followed by PRN monthly injections. The as-needed criteria included an increase of 100 µm in sub-foveal thickness on OCT as well as new macular hemorrhage, new classic choroidal neovascular membrane, or a decrease of 10 letters of best-corrected visual acuity in conjunction with fluid on OCT.

Visual results. Visual acuity for all patients receiving 16 Gy radiation plus two initial ranibizumab injections showed stability at 18 months. Breaking it down between the previously treated and the treatment-naïve cohorts, the previously treated group was essentially stable at 18 months, while the treatment-naïve cohort demonstrated improvement.

Those receiving 24 Gy radiation plus two early ranibizumab injections had a visual acuity gain at 12 months (the longest data lot for that group). There was no appreciable difference between the previously treated and treatment-naïve cohorts.

Patients undergoing the 16 Gy radiation with no loading dose of ranibizumab—a group that included only one treatment-naïve patient—had essentially stable visual acuity at 12 months.

Responder analysis. Both the 12- and 18-month data showed that the overwhelming majority of those receiving 16 Gy radiation and two initial injections achieved the primary safety outcome of fewer that 15 letters lost. Three-line gainers were essentially equal to MARINA at 12 and 18 months.

All of those who underwent 24 Gy followed by two initial injections met the safety criteria of fewer than 15 letters lost. But there were fewer 15-letter gains, probably because of lower baseline visual acuity in this group.

As you might expect, the 16 Gy radiation-first group showed no 15-letter gainers. But their safety profile was good, with an overwhelming majority losing less than 15 letters at six and 12 months, with half gaining visual acuity.

Durability. Can you get rid of injections altogether? Recall that the patients in the first two groups, some of whom underwent 16 Gy radiation and others 24 Gy, all received two baseline ranibizumab injections. Among these, the 16 Gy group received a mean of about two additional injections at 18 months. Twenty percent required no further injections following baseline. A mean of about four injections spanning 18 months indicates good durability.

Those who received 24 Gy displayed a more dramatic response, with a majority needing no additional injections out to 12 months. The group on average received one additional injection over 12 months. Here too we see excellent durability.

The radiation-first group, as you might expect, was more evenly distributed. Using the same ranibizumab retreatment criteria as those of the other groups, an average of fewer than three injections were required to maintain visual acuity.

Anatomic parameters. Anatomic changes play an important part in determining a treatment strategy's durability of effect. All three treatment groups demonstrated a decrease of roughly 75 µm on OCT subfoveal thickness. This effect occurred immediately in patients receiving loading doses of ranibizumab, and by three months in those who did not. This effect remained durable in all three groups.

Fluorescein angiography showed that the choroidal neovascular membrane was shrinking in all treatment groups. Only the total lesion size remained essentially stable for those getting 16 Gy and two baseline ranibizumab injections; all other anatomic parameters were shrinking out to 18 months.

Among those in the 24 Gy radiation group, the total lesion size remained stable out to 12 months. The actual neovascular choroidal membrane regressed for a time but then started to bounce back at 12 months.

Oraya's IRay is an office-based robotically-controlled, externally delivered, stereotactic low-voltage X-ray treatment system for wet AMD that delivers three beams through the inferior pars plana which overlap precisely on the macula.

Even those in the 16 Gy radiation group with no loading dose of ranibizumab with PRN injections thereafter had a durable destructive effect on the choroidal neovascular membrane. This result is something you don't see with current anti-VEGF therapies. That was very exciting.

One patient might qualify as a “poster child” for the success of low-voltage X-ray radiation therapy. Twenty-two months into the study, the patient had never had a ranibizumab injection. Treatment-naïve with good vision at baseline, the patient gained four letters of visual acuity at the 22-month follow-up. Anatomically, there was a reduction in the size of the choroidal neovascular membrane as well as normal foveal contours on OCT. These effects can be attributed solely to radiation.

Safety. None of the patients showed evidence of radiation retinopathy. Nor were there any instances of radiation-induced optic neuropathy or cataract advancement.

Meantime, Oraya has completed enrollment of a multi-center European study that is randomized, double-masked, and sham-controlled. It was designed to evaluate two radiation doses—16 Gy and 24 Gy—with one load of ranibizumab followed by PRN injections vs sham radiation with one load of ranibizumab followed by PRN injections. The primary study extends 12 months with follow-up out to 24 months.


A spirited question-and-answer session followed Dr. Dugel's and Dr. Moshfeghi's presentations at the Retinal Physician Symposium. Excerpts follow:

Question: Dr. Dugel, you noted that patients in your study had received up to 38 injections in the previous five years. That's probably half the number of injections that they should have been given. So they probably came in with a lot of VEGF-driven disease that could be improved substantially because they had been undertreated. This just reflects what's going on, and it's obviously a major issue in the community.

You guys are systematically undertreating patients and then enrolling them into these studies. With 38 injections in five years, these patients on average probably lost more than 10 letters than they should have lost during that period. How did you determine inclusion criteria?

Dr. Dugel: Our study looked at a standing patient population. This wasn't a prospective study from their earliest treatment. It was a preliminary study of 53 difficult patients designed to demonstrate safety as well as biological signals. But as you suggested, it reflects the standard of care in the community. It mirrors realistically what all of us are doing at this point.

Question: The standard of care is severe undertreatment, but we should be striving to do comparative studies vis-à-vis the optimal standard of care. Shouldn't we be trying to come up with alternative approaches that address the treatment burden, but in a way that yields results comparable to the optimal standard of care?

Dr. Moshfeghi: What we're seeing in both the NeoVista and Oraya clinical trials is patients who have been under-treated or treatment-neglected and have the potential for visual acuity gain. This is something we have not seen with other therapies.

Dr. Dugel: Prior to enrollment in the MERITAGE study, patients were getting increasingly frequent injections to about once every four to six weeks. Meanwhile, their visual acuities were decreasing. In all of these study populations, there are patients who received ranibizumab injections and did well and stayed well; there are those who did well and got worse; and there are those who did not do well no matter how many injections they got. The latter two groups are the ones we want to target as people who would benefit from a safe and effective therapy that also relieves their treatment burden.

Question: When you look at the Oraya study's 18-month data, you see visual acuity starting to fall off. What's the current thinking on radiation retreatment?

Dr. Moshfeghi: That's a question that is currently under consideration. It comes down to: Are you too late for retreatment at that point? And should you divide the dose earlier? Beyond that, is 16 Gy or 24 Gy the optimal dose? We saw a very dramatic effect at the 24 Gy dose with greater durability.

Results from two studies should help us answer these questions. One is the phase 1 extension of the 24 Gy radiation group out to 18 months, which should be available in time for the ASRS meeting this fall. The other is the European data.

Question: Patients understandably often have a poor baseline visual acuity, and in one or two cases that you showed, they achieved a five-line visual gain. You would expect that to a certain extent. But I think the eight-letter mean visual gain in your phase 1 data might be difficult to sustain over the long term.

In terms of treating RAP lesions, most people would say these are seen in some 12% to 16% of all wet AMD cases. RAP lesions are not entirely excluded from your studies, but they are seen in a very small number of cases. So there is a role in treating RAP lesions with radiation, but we still need to identify that role. Would you be comparing it with monthly ranibizumab injections, or with these injections in combination with, say, photodynamic treatment?

Dr. Moshfeghi: There's a school of thought that you should do radiation first when dealing with a VEGF inhibitor, which in fact is what they do with systemic cancer. This is because there's a time lag of three to six months before the radiation's mechanism of action kicks in. So what comes first, the horse or the cart?

When evaluating the radiation-first regimen, we enrolled patients with lower baseline visual acuity owing to the stigma that historically has attached to radiation, where we wished to avoid any potential harm from the therapy. At least with ocular disease, concomitant radiation and anti-VEGF therapy around the same time appears to be the best strategy based on the phase 1 radiation-first regimen. We will know soon—based on the European data—how the visual acuity results stack up.

Dr. Dugel: I think there's a subset of patients we've all seen, perhaps 5% to 10%, where you give patients anti-VEGF treatment three to five times and they're done—they'll never return and they're a success. I can see little reason to start out with radiation or combination therapy at first, and most of us don't. But there's surely a place for radiation after a certain number of injections prove ineffective.

Dr. Moshfeghi: Data from our studies and elsewhere have shown that radiation combined with anti-VEGF therapy can cause involution of the choroidal neovascular membrane. This suggests that combination therapy is something we should consider right from the start because it will likely lead to fewer complications.

Dr. Dugel: If what I said about classic lesions is in fact true, radiation might be very beneficial in cases where we see the lesion on fluorescein angiography and say, “This patient I know will respond well to epimacular brachytherapy.” In such cases, we may not wait for four or five injections. We might apply radiation right from the start.

Question: All the studies are designed to do OCTs monthly. So you're really not reducing the treatment burden. I mean, we're still not ready to treat these patients and say, “Come back in six months and we'll recheck you.” So how do you envision this treatment going forward? Do you expect that you'll be able to stop treatment and never follow them? How do you picture reducing the treatment burden overall if we still have to follow them monthly?

Dr. Dugel: You're right: we have to look at the treatment burden as not just the injection but the entire follow-up process. Once you look at something monthly and gain confidence in your data, you can then say, “The data are good enough that once the treatment is implemented, we may not have to look every month. We may be able to look every three or four months.” We hope this happens. But until the data come out and we have confidence in the likely outcomes, we must follow the study design and look at monthly OCTs.

I envision that once we have the data and can identify subtypes of patients that respond best, it will enable us to say, “This patient is done.” At that point, we may not need to see the patient again for several months.

Dr. Moshfeghi: If you look at the 60% of patients in the 24 Gy radiation group who required no injections following baseline and track their FAs, that group is dry. They're not leaking on fluorescein angiography and they're showing no fluid on OCT, because the choroidal neovascular membrane appears to have been destroyed. At what point would you follow up? Could you send that patient out for quarterly assessment?

In our phase 1 trial, we obtained fluorescein angiograms every three months in the first year and every six months in the second. We're closing in on two years for the 16 Gy radiation group. This will provide a greater understanding of what the follow-up will be. If it works, ideally we won't be seeing these patients monthly.

Question: Another group of patients for whom I visualize this having a real impact is those who for various reasons cannot come in regularly. They can't come in monthly; some can't even come in quarterly. It may be access; it may be illness. If you have a therapy that can deliver long-term effects, and they meet certain criteria, that's a niche where I could really see this playing a role. Do you agree?

Dr. Moshfeghi: Unfortunately, in the Mexican data, with both the 16 Gy and 24 Gy doses, there's a subset of patients who continue to leak, and they may or may not correlate with the ability to travel or to come into the practice. Fluorescein angiography may play a dominant role in evaluating patients with combination therapy, in terms of how long you can send the patient out for and comfortably feel that they're not going to recur.

Dr. Dugel: Right now, we cannot answer your question as to where this will fit in. As for the question of whether this is worth pursuing, there's no doubt in my mind that it is.

We've been talking much about treatment burden. We were in a meeting a couple weeks ago where half an education section was about treatment burden. That's a phrase that didn't even exist in our vocabulary a couple years ago. We're finally waking up to reality and recognizing that we should be studying this.

Question: Could you both comment briefly on why you feel that your approach might be better than the other?

Dr. Dugel: Radiation has been applied before in the '80s and '90s. And it didn't stop because it was ineffective; it stopped because there were too many side effects. The beauty of epimacular brachytherapy is that for the first time we can deliver radiation from inside the eye exactly where the target is. In every other study, the radiation source was placed outside the eye. When you think about it, there's a lot more radiation that has to go through a lot more structures if you place something outside the eye.

We know it's effective. The question is whether we can control the safety. And I believe that the best strategy is to use the safest radiation and apply it exactly at the targeted site.

Dr. Moshfeghi: Radiation basically comes down to how you deliver it, where it goes and how tightly you can control that area. If you look at the complications that have been presented from their data, they're not related to the radiation; they're related to the procedure. You get advancement of cataract, you can get hemorrhage, and you can get retinal tears. None of these events happens with an external approach. You don't have a risk of endophthalmitis with external radiation; you don't have a risk of a postoperative wound leak.

When we first presented our data to radiation oncologists and medical physicists, they didn't believe that we could actually have a beam that's 3.5 mm at the sclera and expands to only 4 mm over a distance of 20 mm, because if you do the same thing with a megavoltage beam, at 20 mm your beam has expanded by several fold. RP

A Practice Self-assessment With Activity-based Costing
Pravin Dugel, MD
Even as our volume of anti-VEGF injections has increased, our reimbursement has decreased dramatically, especially this year. Reimbursement this year vs last for OCTs and injections in my practice has dropped 50% and 34%, respectively. This means that for a typical patient getting 10 OCTs and eight injections a year, this year we're reimbursed $1,000 less than last year for the same treatment.
So you open your exam room door and look at a waiting room jam-packed full of people. You look at your watch and say, “Boy, I'm working till eight o'clock tonight. I must be doing well. I must be profitable.”
What we don't always look at is what it takes to bring in those patients and keep them there—the extra office space we had to add and the extra staff we had to hire. So how do we determine profitability in a methodical way?
One of the best ways is a well-known accounting principle called activity-based costing, or ABC. It requires three steps. First, identify your main activities as individual profit/loss centers. Second, assign time and resource values to each activity. Third, determine the dollar-per-unit cost for each activity.
I looked at two years: 2005 and 2007. I picked 2005 because we were still providing traditional retinal services. We were a retinal-surgical practice based mainly on vitrectomies and laser photocoagulation. And I chose 2007 because we had entered the anti-VEGF era. My practice changed in the blink of an eye from predominantly vitrectomies and laser photocoagulation to mostly OCTs and injections. In those two years my top-line revenue soared by 42%. On the surface, it looked like we were doing well.
But consider the operating costs. To accommodate these new services, our operating costs shot up 64%. Consequently, my profit margin actually dropped by 14%. So I was working harder and longer, but not doing as well.
This is where activity-based costing comes in. What's called the contribution margin tells me how I make my money. For every dollar I make, 51 cents comes from laser photocoagulation, 40 cents comes from vitrectomies, and 16 cents from injections. Everything else is a monetary loss.
Another way to view it is the so-called efficiency margin, which gives the return on investment. For every dollar I invest in the practice, I get back 47 cents for laser procedures, 37 cents for vitrectomies and 27 cents for injections. Clearly, my most profitable services are vitrectomies and laser photocoagulation, and these had been overtaken by less profitable activities.
The first time I spoke on activity-based costing, a highly experienced retinal specialist stood up and said, “Pravin, that's absolute nonsense. I make $170 every time I inject.” He was doing so well that he had purchased an adjacent building to construct a parking lot and had bought two additional offices.
Sometimes we fail to understand that there's overhead with everything we do. Calculating that overhead is difficult to do unless you're willing to be disciplined enough to do it in an independent and dispassionate manner. There is a price to accepting risk, just as there is a price to having people who are billers and collectors. All these costs must be figured into your overhead. Otherwise, you're not getting a realistic view of your practice's profitability.