Lecture: Retinoblastoma Care and Research Updates for 2022

During this live webinar, we will discuss the latest advancements in retinoblastoma clinical care and new research findings. Clinical topics will include roles of treatment modalities including systemic, intraarterial and intravitreal chemotherapy, as well as laser, cryo and brachytherapy. We will then talk about new genetic findings and therapeutic prospects in this pediatric eye cancer, and the promise of aqueous sampling for diagnosis and prognosis.

Lecturers: Dr. Timothy Corson, PhD, Eugene and Marilyn Glick Eye Institute & Dr. David Plager, Indiana University School of Medicine, USA


DR PLAGER: All right. Well, good morning. Or afternoon. Or evening. Depending on where you are in this world. My name is David Plager. I am the Professor of Ophthalmology at Indiana University Medical Center in Indianapolis, Indiana, and the director of Pediatric Ophthalmology at Riley Hospital. We have a webinar here where we’re going to talk about retinoblastoma. The first half will be a clinical update, and the second half will be Dr. Tim Corson, who will talk about some exciting things in retinoblastoma research and what’s coming up in the future. So we’ll jump right into it. And talk about a few fun facts about retinoblastoma. Most of you are familiar with these things. But for background, this is the most common intraocular tumor in children. It’s the 6th most common childhood cancer overall. It occurs in about 1 in 15,000 to 20,000 live births, which means in the United States we see about 300 new retinoblastoma patients per year. Worldwide, this translates to about 8,000 to 9,000 per year. We know that these can show up as either bilateral or unilateral tumors. In large part depending on whether they’re inherited or sporadic. This is the first cancer to be associated with a gene mutation. Namely chromosome 13. Now, with all comers, about 5 to 10% of new patients will have their retinoblastoma inherited from their parents. About 25% to 30% will have a new germinal mutation, meaning that it will become inherited from this point onwards, and about 60% of new tumors are sporadic. Know that the most common presentation, at least in this country, is leukocoria, followed by strabismus, and if the tumor is particularly large or longstanding, the patient may present with a red, painful eye. Differential diagnosis is not that long. I think Coat’s disease is perhaps the one that gives people the most trouble, differentiating it from retinoblastoma. Toxocara may be not so difficult, and hopefully we can all distinguish a retinoblastoma from a cataract. Although some of the PFV-type cataracts with the vascularized membrane behind the lens can at first glance be confused with the retinoblastoma. Average age of diagnosis is around a year. Or a year and a half, on average. Younger for the bilateral ones, where they show up earlier. And older, maybe closer to 2 years of age for unilateral. It’s said that about 90% present within the first five years of life. Although in my experience, it’s been closer to 100% show up at that age. Now, it’s interesting to look at who makes the diagnosis. Who first becomes aware that this is a problem. We did a very small study here at Indiana some years ago, where we just looked back at the last 19 patients who came in with sporadic retinoblastoma. No family history, no reason anybody was looking at this child. We wanted to see who made the diagnosis most often. So I’ll ask you. This is an audience participation question. Who do you think most often makes the initial diagnosis? The pediatrician doing a red reflex test, the family doctor, the mother, the father, or some other relative? Please feel free to chime in. Who first detects these retinoblastomas in sporadic cases? I guess we’re gonna give you a few seconds to vote here. All right. Who was the winner? Well, the winner — 60% said the mother. 27% said pediatricians, followed by family doctors. 0% said the father. That’s interesting. Well, the actual answer in our study was in fact 18 of the 19 were diagnosed by a family member. One was actually diagnosed by an MRI. A child underwent an MRI for a totally unrelated reason and was found to have a retinoblastoma. But the important point was that no patients in this study were diagnosed by routine pediatrician red reflex testing. Now, that’s not to say it never happens. It can happen. But by and large, it’s not. And those of you who said it was the mother were exactly correct. The vast majority of these were first detected by the mother. The second most common was grandmother. And those of you who said 0 fathers were also correct. None of these were first diagnosed by the father. No comment. Let’s change our tack here to talk about treatment. Well, in the pre-1800s, there were all kinds of treatments. Diet, leeches, iodide, Venn section, purgatives, antimony, desiccants. Fortunately we don’t use any of those anymore. If we jump to the mid-1900s when we were actually treating these things, the two treatments were enucleation and external beam radiation. Now, enucleation is effective. It curious the disease, but is disfiguring and results in no vision. External beam radiation is better, but lots of side effects we’re pretty familiar with. Cataracts, retinopathy, figure of 8 skull deformity, the effects of radiation on the infant brain, and secondary tumors, both within and without the radiation field. Clearly lots of side effects to external beam radiation. Now, some of you may not have seen these patients with this very characteristic figure of eight head. There still are some around. You’ll see them. These are patients who were treated in the 1960s, ’70s, or ’80s. And the orbital growth retardation caused by the external beam gives this very characteristic facial appearance. At that time, the Reese-Ellsworth classification was in vogue. But this is not really used anymore. The Reese-Ellsworth classification was done when there were no indirect ophthalmoscopes and we were using external beam radiation. So this is of great historical significance, but no longer used. To me, the biggest revolution in the treatment of retinoblastoma came in the mid-1990s when chemotherapy was introduced. Credited mostly to Judith Kingston in the UK, who applied a chemotherapeutic regimen that was used for other small cell tumors to retinoblastoma. Namely VEC. The goal, of course, was to avoid external beam radiation, and it was found that the intravenous chemotherapy was very effective. Very effective for particularly large tumors like this. And macular tumors. It also was effective for reducing other smaller tumors, but these frequently need to be cleaned up with either a laser or cryo, to actually kill off the tumor completely. Let me give you an example of a patient like this. This was a nine month old child who came in with what mom called a hazy right eye. And this is why it looked hazy. There was a huge tumor in the posterior pole of the right eye. Unfortunately there were also tumors in the left eye. These two tumors abutting the optic nerve. And encroaching on the macula. So let me ask you another audience participation question: Right now, how would you treat this child with bilateral retinoblastoma? Or how do you think this child should be treated? Enucleate the worst eye, radiation, intravenous chemotherapy, intra-arterial chemotherapy, which we haven’t talked about yet, or intravitreous chemotherapy. What is the preferred treatment here? All right. Let’s see. Almost 40% voted for intravenous chemo. And a pretty wide assortment of intra-arterial, intravitreal, enucleation, radiation. Okay. Good. All right. Well, this was the year 2000. Intra-arterial chemotherapy didn’t exist yet, and neither did intravitreal. So this child was treated with this intravenous systemic chemotherapy. And here’s what this left eye looked like, after only one round of intravenous chemotherapy. Here is after three rounds, supplemented with some laser to kill off these tumor cells. Here she is after five rounds. And you can see the chemotherapy was extremely effective. The laser also helped. And the nice thing is, at 15 years follow-up, this child has 20/25 vision in this eye. Something we might not have hoped for, originally. The other eye also responded very nicely to the intravenous chemotherapy. Great tumor shrinkage. Great tumor control. The vision is not great in this eye, because of its location. But the point is: Intravenous therapy worked quite well. Now, let’s jump forward about 25 years. Namely, to the present day. This was a very important article published in the British Journal of Ophthalmology in 2020, by the group in Philadelphia, led by Carol Shields. And what this shows is the long-term success rate of intravenous chemotherapy based on international grouping. We haven’t talked about the international grouping. We will in a second. But basically what it shows is: There’s a very strong correlation between grouping, where children with group A have nearly almost 100% success rate — success meaning globe salvage. That is, the eye did not need to be enucleated, and no external beam radiation needed. Close to 100% for group A. 90% for groups B and C. 70-ish or so percent for group D. And about 30% for group E. So what are all these groups? This really is the classification that’s most commonly used these days, when we’re trying to talk about these tumors. Group A are small tumors. Smaller than 3 millimeters. And not near the optic nerve or the fovea. Group B tumors are larger. They can have a small cuff of subretinal fluid. Group C, larger yet. These can have local seeding within 5 millimeters of the tumor. There can be some subretinal fluid, but less than one quadrant. Group D are much more involved tumors. Diffuse vitreous or subretinal seeding. And there can be more subretinal fluid. More than one quadrant. And group E are the really massively involved eyes that are going to have trouble doing well under any treatment regimen. These are eyes with neovascular glaucoma, tumor in the anterior chamber, hemorrhage, et cetera. This is a schematic showing the groupings. Small tumors in group A, slightly larger in B, some fluid in C, and these more involved eyes that are group D, with very large tumors. Diffuse seedings. This is the international classification that we mainly rely on now. So chemotherapy was shown to be very effective, but it is plagued with the same systemic side effects that children get in systemic chemotherapy for any other type of tumor. Lots of them — cytopenia, fever, alopecia, AML can develop, renal toxicity. Now, to be fair, these side effects are relatively mild. At the standard doses of chemotherapy used for these children. But very real side effects. So we jump forward another decade, into the mid-2000s, when Dr. Kaneko from Japan introduced the idea of intra-arterial chemotherapy. Most of you are familiar with this. But it involves threading a catheter through the femoral artery up the circulation to the back of the eye, where a large amount or at least a high concentration of chemotherapy is delivered to the eye with relatively little systemic side effect. In the United States, this was introduced and popularized by David Abramson and subsequently taken up by many retinoblastoma centers throughout the world, really. So typically one, two, or three drugs are delivered. Melphalan, topotecan, carboplatin, or all three of them. It’s been shown that this method delivers about ten times the concentration of drug to the eye, compared to intravenous chemotherapy. It’s typically delivered in two to six monthly cycles, depending on the size of the tumor initially and the response. Typically reserved for children who are at least three months of age or 6 kilograms. They have to be big enough to undergo the procedure. Children that need to be treated at a younger age than this frequently are given systemic chemo, as a bridge, until they’re old enough or large enough to get the intra-arterial. Now, there are significantly less systemic side effects with the intra-arterial, compared to intravenous. But there are local effects. Including death, stroke, or carotid dissection, just from the procedure. But… Fortunately in experienced hands of interventional radiologists who do this kind of thing, these side effects are really very rare. There are, however, local side effects in the eye. You can have vascular occlusion and toxicity to the optic nerve. Particularly with carboplatin. And even other local side effects. Hyperemia, edema — these are annoying, but usually self-limiting. Jumping forward yet another decade, to get us into the early 2010s, intravitreal application of chemotherapy was introduced. This is typically done with either topotecan or melphalan. Dosages in the range of 30 to 60 micrograms for topotecan and 20 to 30 micrograms for melphalan. Injected in very small volumes, 0.6 to 1 milliliters. It’s recommended that the cryo be applied to the site of injection. The idea being if any kind of tumor cells try to escape from the eye through the needle tract, they can be killed with the cryo. Now, these injections usually need to be repeated every two to four weeks, depending on the response. Here’s an example of such a child who has undergone several rounds of intra-arterial chemotherapy and has had a nice response to the main tumor, but you can see in this slide on the left the development of numerous vitreous seeds. So this child was treated with an injection of melphalan, and even one month later you can see the vitreous seeds are essentially gone. There was a second solid retinal tumor here in the green circle, treated with laser, and a month later, it too is a flattened scar. So the melphalan can be quite effective for vitreous seeds, which prior to the introduction of intravitreal chemotherapy were very difficult to eradicate. Here’s another example. This was a child in October of 2019 who had undergone intra-arterial chemotherapy and laser treatment, and these two scars looked quite good. But two months later, they come back with these rather obvious edge recurrences. And the tumor here. You can see these tiny snowballs floating in the vitreous, which are the vitreous seeds. So this child was given two more rounds of intra-arterial chemotherapy. And you can see the intra-arterial chemotherapy was very effective for wiping out the retinal tumors. But you can also see that the vitreous seeds have actually increased in size and number. So at this point, intravitreal injection is indicated. This child was given topotecan. Now, topotecan is not as effective as melphalan. But it’s also more gentle on the retina than melphalan. So you have to weigh the pros and cons of those two modalities. Now, this child needed seven injections of topotecan, given over about a six month period. They don’t usually need quite this many. But by the time August had rolled around, all the vitreous seeds were cleared from the eye. And the other tumors remained good. So intravitreal chemotherapy has been a huge, huge plus in eradicating vitreous seeds. There are unfortunately complications associated with this as well. The idea of local spread through the needle tract is more a theoretic one. It’s really close to 0% worldwide, when these techniques are employed. Retinal toxicity and necrosis, however, is a real side effect. More so again with melphalan than topotecan. And you could have hemorrhage or infection from the injection. This is a child who we inherited. They came from South America. With this sheet giving the history of all the treatment that this child had had previously. Much of which I couldn’t understand. But when they came here, they still had a tumor, a peripheral tumor, that you can see in this eye. We applied cryotherapy to this tumor. And it shrunk the tumor somewhat, but we now have numerous large and medium vitreous seeds. We applied cryo to it again and made plans for intravitreal injection. So the child underwent two intravitreal injections of melphalan, each time showing some nice improvement in the seeds. After the third injection, given on February 21st of 2017, they called back the next day, saying the child says he can’t see. Not the kind of phone call you want to get. The child came in later that day, and you can see some whitening of the retina here. Retinal ischemic appearance. And here two weeks later, even more progression of the retinal ischemia. And a couple weeks later, further progression. So here, several months later, we won the battle. That is, there are no vitreous seeds, and the child’s cured of their retinoblastoma. But we very much lost the war, because we’ve lost the vision, due to the retinal necrosis from the melphalan. So in summary, in 2022, chemotherapy is clearly the mainstay for treatment of retinoblastoma. It can be given intravenously, it can be given intra-arterially, it can be given intravitreally. There is really no unanimity among retinoblastoma experts throughout the world. There are some centers where they use pretty much exclusively intravenous chemo. There are other centers that use almost exclusively intra-arterial chemo. There are some centers that give perhaps intravenous chemo for bilateral tumors, but intra-arterial chemo for unilateral tumors. And there are still others who use some combination of those, depending on the child, the laterality, the size and extent of the tumor, et cetera. No matter which of these chemotherapy regimens you choose, it frequently needs to be supplemented with local therapy in the form of cryotherapy for peripheral lesions, laser treatment, thermal therapy for more posterior lesions, and brachytherapy of iodine plaque for lesions that are too large for these two. Now… People who choose intravenous chemo for most of their patients do so in large part because there is some thought that the intravenous chemo reduces the risk of secondary tumors. Reduces the risk of pineoblastoma, and may reduce the risk of metastatic disease. Not everyone is convinced of these attributes. But this is one of the reasons why intravenous is chosen by some. Regardless of which of these treatment modalities we use, anywhere in the world, the goals of treatment remain the same. The number one thing is to save the life of the child. Because this tumor is universally fatal if not treated. Second is to try to salvage the eye. And third is to try to optimize whatever vision potential this eye may have. Thank you for spending some time with us this morning. Now, we may have some — we’ll have some time at the end to answer questions. We can also answer questions that are put into the chat. If you would like. But we’re gonna go ahead and move on to the second half of this show. I’m gonna stop here. And I’m going to introduce Dr. Corson. Tim Corson is the Meryl Grayson Senior Associate Professor and director of basic and translational research in the Department of Ophthalmology at Indiana University. He holds bachelors, masters, and PhD degrees from the University of Toronto, and completed a postdoctoral fellowship at Yale University, before joining Indiana. He’s been involved in retinoblastoma research for 20 years. Since his doctoral work with Brenda Galley. And he also researches drug discovery for neovascular eye diseases. He’s published over 80 articles and papers, and his lab is supported by grants from the NIH and several foundations. Tim?

DR CORSON: Thank you, Dr. Plager. It’s my pleasure to be here today, and thank you all for joining us. So I’m gonna talk a little bit about some of the most recent advances in retinoblastoma research over the last year or so, given that we are at the beginning of the new year. So I’d like to introduce some of the new genetic findings in retinoblastoma, talk about some developments in new therapeutics for the cancer, and some new experimental models. And finally, touch on a topic that is of interest to a lot of people, the possibility of sampling the aqueous humor for diagnosis and prognosis of the cancer. So let’s go back initially to retinoblastoma genetics, which Dr. Plager introduced. That retinoblastoma comes in two major forms: The sporadic or non-heritable form, and the heritable form. So in nearly all cases, both of these forms are associated with mutations of the RB1 gene, the retinoblastoma gene, which was the first tumor suppressor gene to be identified. The difference between the two forms is that in the heritable form, one mutation, one allele of this gene, is mutated in every cell of the body. Whereas in the sporadic form, this first mutation happens in a single susceptible retinal cell. A second mutation occurs, knocking out both alleles, and therefore all function of the retinoblastoma gene, and given that the heritable patients have this predisposing mutation in all their cells, this happens very frequently. So they tend to develop bilateral multifocal disease. Whereas in the sporadic patients, this just happens in a single cell, so it’s usually a unilateral, unifocal tumor. However, we now know that loss of these two copies of the RB1 gene doesn’t necessarily lead to a full blown retinoblastoma. And instead, leads to a benign tumor called retinoma or retinocytoma. Further events, termed M3 to MN, are necessary for progression to the cancer. And as you can see from this figure, from a review we did a few years ago, there are a number of potential factors that have been identified. And some of these have been quite well characterized. But I should point out that none of these have yet resulted in therapy that is in the clinic. There is also a very small subset of retinoblastoma patients, about 1.5% of unilateral patients, whose cancer is driven not by mutation in the retinoblastoma gene, but instead by amplification of the MYCN oncogene, which is more commonly associated with neuroblastoma and medulloblastoma, but can be the driver of other events and molecular changes that have not yet been characterized. So against that background, there have been a number of attempts to try to classify retinoblastoma based on genetic and genomic changes. And there was one recent one that I just published last year, that was perhaps the most ambitious and exciting of these to date. This was a study from the Institute Curie, and they identified two subtypes of retinoblastoma based on analysis of gene expression across the entire tumor, based on the common patterns of genomic gain and loss that are seen in retinoblastoma, and some changes in DNA methylation. And these two subtypes have some very interesting features. You can see subtype one is more likely to be bilateral. The dark blue color here. More likely to have an earlier age at diagnosis. Compared with subtype two. And more likely to have an exophytic growth pattern, shown by purple, compared with endophytic in yellow. Subtype one is less likely to be necrotic than subtype two. So some very interesting clinical correlates associated with these molecular subtypes. What is perhaps most interesting is that these subtypes may be able to predict metastasis risk. So you can see subtype two stains very strongly, whereas subtype one and normal retina do not. When they looked at expression, the intensity of the level in these two subtypes — you can see a very strong difference — you can see that TFF1 is higher in metastatic cancer, versus large population of non-metastatic cancer. This is still preliminary work, but it does raise the possibility of being able to predicting whether a retinoblastoma will metastasize, based on expression of this marker or other features of these subtypes. So quite exciting new work. I’ll pause for a minute and ask a question: Do you think that this kind of molecular subtyping will influence the care of retinoblastoma in the future? Some possibilities are: Yes. No. It’s gonna be too expensive. No, it’s not necessary for clinical care. No, it’s not gonna reach the point where it’s robust enough for clinical readiness. Or I don’t know. Take a moment to put in your votes, and we’ll see the results in a moment. Okay. A lot of optimism for this kind of molecular subtyping. That’s very exciting to see. So 80% of respondents think that this will influence clinical care. A minority, about 6%, think it won’t be necessary. 10% don’t know, and a couple of percent think it might be too expensive or not reach clinical readiness. So that’s very reassuring and exciting, that this work has value. Let’s move on now to talking about discovery of novel therapies for retinoblastoma. So as of today, there’s still no targeted therapies for retinoblastoma that are widely clinically used. This contrasts with other cancer types. So the Food and Drug Administration in the United States has approved more than 80 different targeted drugs for 33 different other cancer types. So clearly retinoblastoma has some room to develop in this area. And there are a number of clinical trials underway for retinoblastoma. Many of them are looking at new delivery options for existing chemotherapies. Some of them are as part of larger trials, enrolling a variety of childhood cancers. But I just wanted to highlight some of the current retinoblastoma-specific trials that are ongoing. A couple involving antiangiogenic drugs, conbercept and ranibizumab, one with Visudyne, one with temozolomide, and one interesting one with an oncolytic virus called VCN-01. So we await the results of those trials, but there is work going on targeting preclinical studies. I would like to highlight one with this illustration. This is one of the first preclinical studies showing adoption of immunotherapy for retinoblastoma. This is from a group at the University of North Carolina. So they employed the CAR-T-cell therapeutic approach, chimeric antigen receptor T-cells. For those not familiar with this approach, T-cells are extracted from the blood of the patient, engineered to express a T-cell receptor that is able to recognize the tumor, they’re then administered back to the patient, and those T-cells can target the tumor, and the patient’s own immune system will eradicate that tumor. So in this study, the authors took advantage of a protein called GD2, which is expressed on the surface of retinoblastoma cells. It’s also seen on the surface of neuroblastoma cells, which is where this idea came from, since it’s been successfully targeted in neuroblastoma. They also engineered the CAR-T cells to express interleukin 15, which stimulates T-cell function, and embedded the cells in a hydrogel. So this gel allowed enhanced delivery when they applied it to eyes of mice that had had human tumor cells injected into them. And what you can see here — this is the tumor growth on the y axis of these graphs. If you look at the green curves, this was with their GD2 targeting CAR-T-cells with IL-15, and it basically completely blocked tumor growth. Far better than a control condition or components of the therapy administered separately. What’s even more exciting, this CAR-T-cell treatment was essentially able to lead to 100% survival in the treated mice, compared with the control condition, where all of the mice died or had to be sacrificed within about 40 days. So I think this is a very promising preclinical study that paves the way for therapy for retinoblastoma. We also need new models for therapeutic testing in this cancer. For testing new agents, new delivery strategies, evaluating new targets. Most of the preclinical studies, including the one that I just presented, revolve around injecting a few different human cell lines, many of which have been available for decades, although there are patient derived xenograft models, where tumors are taken from patients, grown in culture, and implanted into animals, which captures more of the variability of the human disease. In terms of developing new models, one of the big innovations over the last year was the use of organoids for modeling retinoblastoma. This is work from Wen Ju Medical University in Beijing. Organoids are stem cell derived clusters of cells that recapitulates some of the features of the organ that they are isolated from. And there’s been a lot of work on retinal organoids across the spectrum of retinal disease. So in this study, they took human embryonic stem cells, and they genetically engineered them using CRISPR Cas9 gene editing to mutate or knock out the gene, they engineered human retinal tissues, and grew clusters of cells. I should point out that St. Jude’s Children’s Hospital did similar things with patient derived pluripotent stem cells, cells derived from actual patients, rather than being gene edited, which perhaps has more potential for personalized medicine in future. So if you grow these organoids over quite a long period, 120 days, you can see the ones with mutated RB1 in the top or knockout RB1 in the middle row form this edge which is quite disorganized, irregular, quite different from those seen in the wild type organoids, and these, the authors went on to show, have the properties of retinoblastoma tumor cells. They histologically look like retinoblastoma, and have many of the genetic changes seen in retinoblastoma. So this creates a really great cell culture-based system for testing new therapies in future. Looking at animal models now, this is work from the lab of my collaborator Anthony Daniels at Vanderbilt University. They’ve done animal models, rabbits for intra-arterial and more recently intravitreal chemotherapy, which as Dr. Plager mentioned is one of the really hot areas in retinoblastoma therapy right now. So by introducing a human retinoblastoma cell line intravitreally in rabbits they can see these vitreous seed-like structures in the eye of the rabbit. And here they were testing a drug called Belinostat, a histone deacetylase inhibitor, based on the fact that they’re dysregulated in retinoblastoma. You can see that the intravitreal injection given over multiple courses really eradicated these seeds. Quantified here. And when they looked histologically at these eyes, here is the untreated eye with a number of seeds above the retina, and very few — really just single cells in the Belinostat treatment. Quite similar to intravitreal melphalan, which is the most widely used drug right now. So in the final few minutes, I want to talk about aqueous humor sampling. This is a new potential way of diagnosing and prognosticating retinoblastoma. It’s been done largely by a group at UCLA and University of Southern California. Also a group in Birmingham, UK, working on this. Basically the aqueous humor is an accessible source of tumor material from intact eyes. As more and more eyes are saved, rather than enucleated, it’s becoming increasingly difficult to obtain tumor material for research and also for prognostication of the patients. So as you can see in this really excellent figure from a recent review, the tumor in the back of the eye is throwing off cells into the vitreous, and tumor derived cell-free DNA as well. Some of this finds its way into the anterior chamber, into the aqueous, where it can be sampled by paracentesis, taking out about 100 microliters of aqueous with a 32-gauge needle, that is routinely done during intravitreal injection to normalize intraocular pressure. So sampling aqueous has the potential for diagnosis. Maybe for subtyping in future. And prognosis of the cancer. The circulating tumor DNA — so there’s the fraction of the DNA extracted in the aqueous, or tumor fraction, TFX, decreases during treatment, but the copy number changes that are distinctive for retinoblastoma remain quite consistent. And just to show you what I mean, here is a case study from a recent paper. Here’s the right eye of a patient that was a group C eye. Stage C2b eye. That received intravitreal melphalan. So at the time of diagnosis, the aqueous humor showed a nearly 80% tumor fraction. So a lot of the DNA in the aqueous humor was tumor DNA. And you can see some of the genomic changes that are commonly seen in retinoblastoma. Gain of chromosome 1, gain of chromosome 6, loss of chromosome 16. As this tumor went under a couple of rounds of intravitreal chemotherapy, not only did the tumor fraction decrease from 60 down to just 5%, but the prevalence of these changes decreased as well, indicating that the tumor cells in the aqueous were being reduced. And this eye was actually successfully treated with the intravitreal melphalan. In comparison, the left eye of the same patient, which was a group D eye, underwent several rounds of melphalan and topotecan, and you can see it didn’t improve, and at the time of enucleation, the tumor fractions were even higher than diagnosis. The genetic fractions were evident. And you can see changes very similar to what was seen in the aqueous humor. Further evidence that the aqueous humor biopsy reflects what’s seen in the tumor. So does this approach have potential for prognostication as well? A number of studies have suggested that yes, it does. Here’s sampling of aqueous humor and looking at the ocular survival of tumors with gain of chromosome 6p. Versus without gain of chromosome 6p. So this is genomic change that for decades has been known to be important in retinoblastoma. But this was the first study sampling this in the aqueous humor, and showing that ocular survival was worse in patients with this gain in their aqueous humor biopsy. So I’ll pause and ask another question similar to the previous one. Do you think that the aqueous humor biopsy will influence clinical care in the future? Possibilities are yes, no, it’s too expensive, no, it’s not necessary, no, it won’t be robust enough to reach clinical readiness, or I don’t know. Okay. Similar results to the last question. 85% of people think that this will influence clinical care in future. A couple of percent each think it might be too expensive or not necessary. Or won’t reach clinical readiness. And only 7% don’t have an opinion at this time. So I think this is definitely an area to watch. There are a number of studies coming out on this topic. And I think we can expect to see a lot more in future. So in the last couple of minutes, I would just like to offer my thoughts on where things are going this year and in future. It’s an exciting time for retinoblastoma research. I think we’re gonna see more integrative “omics”-type analyses. By this I mean proteomics, genomics, transcriptomics, metabolomics, looking at broad features of retinoblastoma tumors. Perhaps from aqueous humor biopsies. Just as one quick example here, this is a very recent study from the group at Shanghai. Looking at a panel of metabolites. Several different metabolites that were detected in the aqueous humor. And this panel had very good predictive power for detecting progression of retinoblastoma. So I think we’re likely to see more studies like this in future. We’re gonna continue to see identification of new target proteins and new therapeutic approaches. And I think it’s likely that we’ll start to see more subtyping from aqueous samples. If we can apply that new subtyping approach that I mentioned to aqueous, we can get a lot of data about where the tumor might be going, clinically. Obviously the new therapies can be tested in different systems. For instance, Belinostat in organoids, other therapies in intravitreous and intra-arterial chemotherapy models. And I also think there’s a lot of room for imaging-based research. So as more and more eyes are saved, in order to figure out what’s going on with the tumors, what are the clinical features of those tumors, I think we can turn to radiomics, which is using machine learning to identify features of MRI scans, for instance, that could subtype tumors further. And there’s already been some interesting work in this direction. Finally I would like to close with I think a very positive slide. This is from a recent bibliometric study looking at the cumulative publications related to retinoblastoma. As you can see, over the last 20 years, it’s been exponentially growing, and if that continues, we have a lot to look forward to in retinoblastoma-related research in future. I would like to thank members of my lab and my collaborators. In bold are my retinoblastoma-related collaborators, and our various funding sources. But I will close and leave up on the slide for a few minutes just a list of references. These are those I talked about today and a few others from the last year that I didn’t have time to talk about, but are very exciting as well. If you would like to take a screenshot of that, feel free. Other than that, it’s time to open up the floor for questions for either Dr. Plager or myself. Thank you.

>> Thank you, Dr. Corson. There are a few questions in the Q and A. If you or Dr. Plager would like to go through them.

DR CORSON: Okay. The first question I see asks: Why is subtype named RB635? So this relates to that molecular subtyping study that I mentioned. Those were just two different tumor samples that they were highlighting in the paper. It was just their internal code for that particular case. Another question. Does AC sampling detect the incidence of recurrence of tumors? To my knowledge, we haven’t seen any studies showing that yet. This is where I think there’s a lot of excitement around potentially looking at kinds of subtypes and prognostic factors in the aqueous humor that could be predictive of tumor recurrence or future metastasis. But I don’t think we’re quite there yet.

>> Somebody wants your email.

DR CORSON: Okay. In regards to the CAR-T immunotherapy, is there any research regarding MR1 and retinoblastoma? I am not familiar with any. It doesn’t mean it’s not out there. But I haven’t seen anything. If anyone in the audience knows about that, I would be happy to hear in the chat or in the Q and A. Are there some studies in Asia about aqueous humor therapy? I’m sure they’re ongoing. I think this is an area that has captured the attention of a lot of people worldwide. So I think it’s something to be looking out for in future. Okay. A couple of questions that maybe Dr. Plager is better equipped to ask. Can the AC sampling be done without concurrent intravitreal chemotherapy?

DR PLAGER: Yes, I’m sure it can. AC sampling is really not part of routine clinical practice right now. But it certainly is being done under research programs.

DR CORSON: Can we perform the AH sampling in every stage of RB? I would say that up to group D it’s probably possible. Group EI with extensive extraocular involvement. It might be more challenging. Dr. Plager, I welcome your thoughts on that.

DR PLAGER: Yeah, well, traditionally, we’ve always been very hesitant to put any needles in an eye that had retinoblastoma. We’ve sort of overcome that in the last ten years or so, with the intravitreal chemo. But you still think a little bit about it, when you’re sticking the needle in an eye with active tumor. And certainly you think about it multiple times before you stick it in the anterior chamber of an eye that had anterior chamber tumor.

DR CORSON: The next question relates to that. Will AC sampling not increase the risk of converting intraocular tumor to extraocular tumor?

DR PLAGER: Yeah. Of course, that’s the theoretic risk I was just talking about. It turns out this risk, particularly with the intravitreal injections, has been basically none. I don’t want to say it’s none. But it’s been basically none. But that includes putting cryotherapy on the site. Now, the anterior chamber sampling is a little different. But again, it’s something you think about. But it turns out in reality it has not to my knowledge been known to have happened. Yeah. There’s a question here about differentiating Coat’s disease from retinoblastoma. Again, I don’t think I would say that anterior chamber sampling is clinically applicable to that at this point. There really are other ways to differentiate the two. And I think an experienced observer can tell the difference by looking at it, and also from doing non-invasive studies. Either neuroimaging or ultrasound. They can be differentiated without doing the sampling.

DR CORSON: What about transscleral drug delivery rather than intravitreal injections? So there are a number of studies ongoing, looking at different delivery options. For cytotoxic chemotherapy, for retinoblastoma, including transscleral, subconjunctival, there’s a role for plaque delivery as well. So there’s a lot of things in the works. I haven’t seen too much in the literature yet. Dr. Plager, I don’t know about you.

DR PLAGER: There is a lot of work about this kind of alternate ways of drug delivery. Even putting — there are clinically applicable, for instance, people are putting steroids into of the nasolacrimal duct after cataract surgery, and it gets into the eye very nicely. So those kinds of alternate drug deliveries I think are going to become a real thing. I don’t think they are just yet for retinoblastoma.

DR CORSON: What about the potential role of blood-based liquid biopsy for RB diagnosis and prognostication? So of course, in the case of a patient with heritable disease, there will be a detectable RB1 mutation in their blood in nearly all cases. So in that special case, it’s possible to do genetic testing of their offspring. To see if they are at risk of the cancer. But I think more broadly, looking at all RB cases, it is at least theoretically possible for a blood-based liquid biopsy. This is obviously something that has garnered a lot of interest in other cancer types. I think the thinking is that the aqueous humor biopsy, although it’s slightly more invasive, does get more tumor DNA closer to the side of the primary tumor. Easier to distinguish that tumor DNA from the bulk of normal DNA. Which is why it’s become so popular.

DR PLAGER: Is there a mechanism of action other than chemotherapeutic agent for intravitreal injection in RB? I don’t believe there is clinically today. That’s not to say that some of these modalities that Tim has talked about probably will be in the future. Other targeted agents. But not for right now. Looks like you got them all, Tim.

DR CORSON: Likewise.

>> All right. We can wait about one minute and see if any more questions come in. If not, I think we can end the session.

DR CORSON: We do have another question.

DR PLAGER: I think technically, primary retinoblastoma has to be in the retina. But there are of course cases of what we call trilateral retinoblastoma, where the tumor is in the pineal gland. And there are cases with just the pineal gland tumor without the retinoblastoma. But other than that, it needs to be in the retinal cell.

DR CORSON: But of course, patients who have a germline mutation, the mutation in the RB1 gene in all their cells, are also at risk for other primary cancers later in life, such as osteosarcoma, other soft tissue cancers, and melanoma as well.

DR PLAGER: Yeah. There’s another question here about the intravitreal chemo. How often do you do it and when do you stop? The treatments can be given every couple of weeks. I tend to do it sort of monthly. So we have enough time to see what the response is. And you really gauge the additional treatment based on what the response is. So as long as you’re making progress, you can continue to do the injections. That case I showed where we did 7, I don’t recall doing more than that on anybody. And with melphalan, that usually takes even less. But at some point, I think particularly if you’re trying to use topotecan, and you’re not getting the kind of response you would like, it makes sense to switch to melphalan. Because it is more effective. But again, it is also more toxic. So it’s gauged on the clinical response. The first question there I think is more a comment than a question. And that is the theoretic — the ethics of trying to salvage stage E cases. And I think a lot of people would agree with what is typed there. That maybe trying to salvage some of these cases is not worthwhile. I wouldn’t make a blanket statement about that. But I’m certainly aware of some cases where the position has been… I think what we would all agree… Overly aggressive in trying to salvage eyes that really had no case to be salvaged. And they ended up losing the child. So you’ve got to keep the big picture in mind. And a lot of these stage E cases are best treated with enucleation. Talking about a 48-gauge microneedle using melphalan to inject to the central retinal artery. I think that would be sort of technically challenging. And I think just using the intra-arterial method with the catheterization gets the medicine where you want it. I’m not sure about using the needle to directly inject the central retinal artery. We don’t currently have any training programs on ocular oncology here at Indiana. There are some at some of the other centers around the world. Is there any role for vitrectomy in retinoblastoma? I would say absolutely not. No. There again have been cases where people have done vitrectomies for retinoblastoma, not knowing that it was a retinoblastoma. Thinking it was something else. And those children have not done well. So that is not to be encouraged.

DR CORSON: Any role of CSF liquid biopsy for detecting CNS mets for extraocular tumor especially where cytology is negative. There have been a number of research studies in this area, using molecular methods to detect either the RB1 mutation of a tumor, or other retinoblastoma features. To monitor minimal residual disease in the case of metastasis. So yes, this has been done on sort of an individualized basis. And there are some references going back a few years on that.

DR PLAGER: Yes. Certainly spinal taps are frequently used in children where there’s concern about CNS involvement.

DR CORSON: Any role for anti-VEGF in treatment? So this is a very interesting question. Quite close to my heart, since my lab also works on angiogenesis. There have been some preclinical studies looking at antiangiogenic approaches in retinoblastoma models. And one small clinical study that I’m aware of, looking at bevacizumab. And as I mentioned, there are a couple of clinical trials with antiangiogenic agents underway in retinoblastoma. But nothing conclusive, as of yet.

DR PLAGER: Question about intra-arterial chemo, if it did not give control after two sessions, do you continue trial? I think it depends on what agents are being used. Some people use a single agent. Some two, some three. For the more involved tumors, we’ve gone sort of directly to using three agents. So if after a couple of sessions of the maximum therapy you’re willing to give you’re not seeing control, then certainly you need to think about something else. But I don’t think if the patient has failed a full trial of three drug intra-arterial, I don’t think intravenous is likely to be helpful, because the drug dose is much higher with intra-arterial than it is intravitreal. You’re gonna have to go to something else. Perhaps something including plaque. But presumably, if the intra-arterial hasn’t worked, the plaque is probably not gonna be able to treat the entire tumor. You would really have to go to even considering external beam radiation, which we pretty much never do anymore. But theoretically, that could cure the tumor, but it’s gonna cause so many side effects, most people won’t go to that. Do you recommend intravitreal or intra-arterial chemo for bilateral retinoblastoma? Well, the intravitreal is mainly reserved for seeds. The basic tumor is usually treated with either intra-arterial or intravenous. And as I mentioned, there’s not unanimity on this. Some people will do only intra-arterial. Some will do only intravitreal. Excuse me. Intravenous. And some will do either. I’m in the group of doing either. We’ve gone much more toward doing intra-arterial, even for bilaterals. Initially, we didn’t do intra-arterial for bilaterals. We did intravenous. But the intra-arterial has been so successful with so few side effects — it really has become our go-to for even bilateral cases. But that’s not a universally held opinion.

>> All right. That looks like all of the questions, and we’re right at the top of the hour, so I think we can go ahead and end here. So I just wanted to thank you both for your time.

DR PLAGER: Thank you.

DR CORSON: Thank you!

Download Slides (Dr. Plager)


Download Slides (Dr. Corson)


January 12, 2022

Last Updated: October 31, 2022

1 thought on “Lecture: Retinoblastoma Care and Research Updates for 2022”

  1. Unfortunately, It is hard to discuss the advanced treatment of intra-arterial and intra-vitreal chemotherapy within the setup of limited resource countries. A viable alternative is to discuss what and when to enucleate vs. applying radiation and intravenous chemotherapy.


Leave a Comment