Predicting Survival using Genomic Adjusted Radiation Dose (GARD) for HPV-Positive Oropharyngeal Squamous Cell Carcinoma (OPSCC)

Podcast Transcript

Dale Shepard, MD, PhD: Cancer Advances, a Cleveland Clinic podcast for medical professionals exploring the latest innovative research and clinical advances in the field of oncology. Thank you for joining us for another episode of Cancer Advances. I'm your host, Dr. Dale Shepard, a medical oncologist here at Cleveland Clinic overseeing our Taussig Phase I and Sarcoma Programs. Today I'm happy to be joined by Dr. Jake Scott, a radiation oncologist and an associate professor and staff physician scientist at Cleveland Clinic. Dr. Scott is here today to discuss genome-based radiotherapy dosing and predicting overall survival and patients with HPV positive oropharyngeal squamous cell carcinoma treated with radiotherapy. So welcome, Jake.

Jacob Scott, MD, DPhil: Thanks, Dale.

Dale Shepard, MD, PhD: Maybe to start off, give us an idea what is it you do here at Cleveland Clinic?

Jacob Scott, MD, DPhil: Great question. So, in addition to being one of your colleagues in the sarcoma program where I take care of patients who have sarcoma using radiation therapy, most of my time is spent as a physician scientist. So that means I run a laboratory. We're primarily NIH-funded, and we study, I'd say quite broadly study why the good therapies we have stop working. That's the big picture. And so that covers everything from radiation therapy, which is my clinical expertise, and why that doesn't work when sometimes it does, and in what situations does it not work, and how we can ameliorate that, all the way through targeted therapies, chemotherapies. And actually, we even study antibiotics in the lab.

So as a species, we, humankind, have done an amazing job of understanding how to combat these bad diseases like cancer and infections. We have these amazing drugs with these amazing protocols, and they work for a lot of people. And sometimes, they don't. And so, what we really focus on is when they don't and why they don't, and how to fix that.

Dale Shepard, MD, PhD: Very good. And so, we're going to be talking about a study that was presented at the American Society for Radiation Oncology annual meeting. Tell us a little bit about the background for that study that you did.

Jacob Scott, MD, DPhil: Yeah. Sure. So yeah, I think that the baseline for this study is the concept within radiation oncology. I just gave you a broad picture of my own research, but we're going to focus in a little for the purpose of this conversation on radiation therapy. So, radiation therapy, for those listening, is the most common agent in all of cancer. So, if you have cancer, the most likely thing that you'll have is radiation therapy across all diseases. Something like half of all cancer cures have a basis of radiation in the backbone of their care. Now that's maybe radiation alone. It can be radiation together with chemotherapy. It could be surgery, and then radiation. But in some way, radiation is responsible, at least in part, for half of all cancer cures.

And so, for the study we're talking about today, we're focusing in on when radiation works and when it doesn't, when it works better, when it works worse, how to figure that out. And in particular, the abstract that we're going to discuss is focusing in on a specific site, so head and neck cancers, and we're talking about the oropharynx, so in the head and neck world. And we're talking about one of the two subsets. In head and neck cancer, there's broadly HPV positive and HPV negative. I think likely that our results that we're going to discuss today apply to both. But the specific results we'll talk about today are in the HPV positive realm.

And what's interesting about HPV positive oropharynx cancer, which is the standard of care would be chemotherapy plus radiation, is that these patients actually have very good outcomes, which is a good thing. So, the question now, the standard of care or the question of the day is for whom can we reduce therapy? Can we give less chemo? Can we give no chemotherapy? Can we give less radiation therapy? For whom can we reduce care and get equivalent outcomes? That's where we are in the standard of care. And there's many clinical trials looking at reduced doses of radiation, like I said, omitting chemo. And what we're looking at is trying to find a genomic basis, so specifically a tumor biopsy, and then a genomic test, and then looking at the data from that to determine which patients would do as well with less therapy compared to standard therapy.

Dale Shepard, MD, PhD: And I guess just to define, when you say less therapy, are you typically thinking about fewer fractions of radiation? Are you thinking about less total energy? What does that mean by less?

Jacob Scott, MD, DPhil: Great question. So, in radiation oncology, as you hinted at, we have different ways of turning up and down the volume, if you will. And nowadays especially, we're giving higher doses for fewer days, something called hypofractionation. But what we're looking at in this study in particular is actually just less or fewer fractions. So, I think we know less about the radiobiology of what we call hypofractionation, which is the five-day or one-week courses. But we know a lot more about the five- and six-week courses that are more standard. And I would say in head and neck cancer, seven weeks is the standard, and we're moving down toward fewer for some patients, like an HPV positive, for example, down towards six weeks. But really, any physician who's practiced oncology knows that every patient responds differently. You give the same dose of chemo to two patients; they respond differently both in their toxicity and in their tumor response.

And so, what we're looking for here is a genomic basis of that response to radiation therapy. And in particular, we're looking at a dataset, a prospectively collected dataset from a European collaborative group in this study of several hundred patients who are treated with HPV positive oropharynx cancer with radiation, most of whom are treated with radiation and chemo as definitive therapy, and that all those patients had tumor biopsies that were sent out for genomic testing. And in this case, that means expression arrays on something called an Affymetrix chip for the techies in the audience. And what we look at is those tens of thousands of genes and try to really figure out from that genomic information, can we tell which patients will do better or worse with less or more therapy?

Dale Shepard, MD, PhD: And so, what have been preliminary results so far?

Jacob Scott, MD, DPhil: So, what we reported at ASTRO and we're writing up as we speak is actually quite a strong signal for a genomic test that's been developed called the radiation sensitivity index. So, it's a genomic test for those familiar with things like Oncotype or MammaPrint in breast cancer or Decipher in prostate cancer. What this is a score, a genomic score based on a tumor biopsy, just as all the others. It really tells you on a scale of zero to one how sensitive or resistant a patient is to radiation. And what's nice about radiation therapy, which is one of the reasons I like the field, is it's very mathematically driven, is that we actually have a series of equations that help us figure out from a sensitivity score how effective a given dose of radiation will be. So, like I said, any Rad Onc knows that 50 gray given to one patient, which gray are the units of our radiation therapy. So, 50 gray of radiation given to one patient might affect a different patient differently or almost always will.

But we really have no way of determining which patient will experience more or less effect from the same dose. And what this radiation sensitivity index does is allows us to calculate what we call the genomic-adjusted radiation dose, which is part of the title of this talk, which is GARD. And so, what we're really suggesting is that we should think about dosing patients toward the effect we want rather than a standard dose. So instead of saying, "I'm going to give 10 doses of radiation to this patient," I'm going to say, "I want to dose them until I have the desired effect." And that's where we're moving with this.

And the preliminary results show us that we're able to very beautifully discriminate within this cohort of 200-something patients who will do better than whom based on the effect that we calculate. And it's not based on necessarily the dose they get. So, it turns out some patients who have a low dose given have a high effect because their tumors are very sensitive, for example. And on the other hand, there's some patients who are given a higher dose but because of a resistant tumor have a smaller effect. And really, it's based on that effect that the outcomes depend not on the dose itself.

Dale Shepard, MD, PhD: And I guess just to clarify, when you say treat to an effect, is this to a predicted effect based on the score, not by serial measurements of something like a genomic signature after therapy?

Jacob Scott, MD, DPhil: That's exactly right. So, in the new paradigm that we're suggesting here, you would take a tumor biopsy, you would calculate a genomic score based on a genomic test, and from that, mathematically, you could predict the effect of a given dose. I have a patient, they come in, they get score A, I give them dose B, I can now predict A times B equals effect. And then it turns out that that effect in this cohort is really what helps us determine outcomes statistically. And so, the hope there is that in the future, instead of saying, "Oh. Well, you have tumor type X, in this case HPV positive oropharyngeal cancers. You need dose Y," we would instead say, "You have tumor type X and a genomic score of Y. You need a dose that is modulated by that."

Dale Shepard, MD, PhD: So, this particular study looking at HPV positive had neck cancers and that's what happened to be looked at, and you alluded earlier that probably doesn't matter about HPV status. Is that based on the genomic changes or just the underlying biology of the tumor?

Jacob Scott, MD, DPhil: So, both. And I think what's interesting about this study is it's literally the largest cohort that we've studied in a single go because the European group that we're collaborating with really did a great job of collecting this tumor type, doing the genomic testing, collecting the data appropriately. But before this study, we've actually published a couple of thousand patients worth of pooled data from lots of different disease sites. And it turns out that the same approach really works every time we test it. So, we've looked at breast cancer, brain cancer, endometrial cancer, muscle cancers like you and I treat, every tumor type that we've looked at with the exception of liquid tumors. So, every solid tumor type we've looked at, this approach of genomic dosing works better than empiric dosing. And so yes, today we're talking about HPV positive oropharynx cancer. But in every single solid tumor we've tested, we think that the same approach works. And so, what we're hoping to move toward is a paradigm where radiation dose is determined by individual patient tumor genomics rather than more broadly by disease type.

Dale Shepard, MD, PhD: And so, I guess, what do you see as the barriers to this becoming more of a mainstream approach?

Jacob Scott, MD, DPhil: That's a great question. I mean, I think that a good problem in radiation oncology, we have excellent outcomes. For the most part, in Rad Onc, we do very well. Patients experience therapy with minimal side effects and our cure rates are very good. And so, for us to adopt a new paradigm where we change the way we dose to tumors is a risk, to be fair. I think it's a little bit less risky from the physician side to increase dose. If I predict that I'm not giving you enough, that's one thing. But it's scary to say, "Hey. You have a very curable cancer, but I think you need less therapy." That's scary as a physician, and I imagine as a patient as well. If I'm going to quote someone a 90 percent cure rate with a standard approach, and then say, "Hey. I have a novel approach where I think you only need half as much therapy, but I'm not sure. I might really reduce your cure rates," that's scary.

And I think we've reached this excellent local optimum where it is difficult to take risks because we're in situations where we have great outcomes now. And obviously, we'd like to improve them. And improving outcomes is a two-way street. So, you can improve outcomes either by increasing cure rates or by keeping cure rates the same and reducing toxicity. Both of those things are, I would consider it improved outcomes. But from the perspective of both the physician and the patient, it's scary to think about dose reduction in this setting even though we know, HPV positive oropharynx cancer is a great example, we are already doing dose reduction studies in this subset because we know they need less dose than their HPV negative cohort.

Dale Shepard, MD, PhD: Or example you used before where instead of five weeks of treatment, you get five days of treatment. There are still patients who might give pause and think more is better.

Jacob Scott, MD, DPhil: Absolutely. And that's absolutely right. And if I was a patient, I'm not sure if I would be interested right now in dose reduction. But there are certainly active trials now even outside of what we're talking about today for dose reduction in this cohort for HPV positive cancers. It seems as though less might be as much or as good for many of these folks as more is.

Dale Shepard, MD, PhD: This sort of approach, simply from the testing standpoint, what does it add in terms of cost and time?

Jacob Scott, MD, DPhil: I would say that the answer to both of those depends a lot on what I would call market forces. So, if you're running a bespoke test individually for a single patient, it's going to take a long time and cost a lot. It's an economy of scale. However, once the test becomes widely available and there's large-scale industrial partners that do it, the costs come down and the time comes down. So, I'd say right now, this is actually not yet a commercially available test. It is available for research studies through Moffitt Cancer Center, but it's moving toward commercially availability. And I think that it's basically going to be based on adoption as people desire to know this information.

So, I think it's really analogous to HPV status itself. So, if you rewind the clock 15 years or so in head and neck cancer, we didn't know about HPV positive status. So, we didn't know that there was such a thing as HPV positive versus HPV negative oropharynx cancers. In seminal papers written in the 2000s, we started realizing that there were these two buckets, and really, the HPV positive cohort did incredibly markedly different 20 percent, 30 percent, 40 percent differences in survival at five years. Today, you cannot treat a patient with head and neck cancer without knowing their HPV status, but you don't have to change your treatment if you don't want to. So, you have to know prognostically which one they are, positive or negative, but you have a choice as to whether to change your treatment paradigm based on information or not. I would argue that this radiation sensitivity index and genomic-adjusted dose paradigm that we're suggesting here is similar in that in the study we show here, the statistics are quite strong. The signal is as strong as HPV status or close to it.

And so therefore, I would argue that really, we shouldn't be treating patients without knowing this information when it comes to prognostic ability, stratification, discussion with our patient about how they're going to do. But it isn't necessary that we changed our strategy quite yet. However, we have a paradigm in which you could change the strategy. So, I think that we're reaching a point where at least in certain disease sites today, the data's strong enough, the one I present here, which of course, is pre-peer review, but we'll get there. But I would argue that we're getting to the point where this is information that really, we should have as radiation oncologists when we're making our treatment decisions and having our discussions with patients.

Dale Shepard, MD, PhD: And I guess just from a perspective of scale, and when we think about differing doses, and you had given an example of somebody getting half the dose, but practically speaking, what is the range? I mean, in terms of either fractions or amount of the dose, how much variation do you see?

Jacob Scott, MD, DPhil: So right now, the standard of care for head and neck cancer is something like seven weeks. So, 70 gray and 35 fractions is kind of the bog standard and appropriate and clinical trial tested and vetted dose. There are active clinical trials that have reduced that dose of 70 gray to 60 for HPV positive patients because it seems as though in retrospective studies that less dose is equivalent. Further, people have reduced it from 70 to 60 and have also emitted chemotherapy. So, a lot of times, most patients with oropharynx cancer get a combination of chemotherapy and radiation at the same time. But in these HPV positive patients, the dose reduction has taken the form of either fewer fractions, so usually 5 fewer fractions or 10 less gray, or a reduction in chemo or removing chemo. And so, I would say in the study that we're talking about in particular, the dose range is somewhere between 45 and 70, where actually I think even 75, there were some patients who received more. And so, there's actually quite a large range.

And for folks who treat this disease, the difference is big in toxicities. 45 gray given, patients will experience almost no toxicity, almost no changes in swallowing function, salivary function, a lot of the other side effects.

Dale Shepard, MD, PhD: Skin effects.

Jacob Scott, MD, DPhil: Skin effects. Whereas the higher doses of 70 and 75 gray, patients can definitely have both short- and long-term effects. And so being able to reduce from 70 to 60 or 50, if we could do that in what we call isocurative way, meaning in a way that we don't change the cure rates, if we could do that, we would significantly be able to reduce toxicity. So, I think it's exciting even in the range that we're already seeing. In particular in this study, we saw patients who received as few as 45, I think gray, and as much as 75.

Dale Shepard, MD, PhD: So, I guess again, this study is looking particularly at HPV positive oropharyngeal cancer. Just broadly speaking, as a radiation oncologist, is there a particular setting, particular tumor type, particular setting that you think this has the potential to have the greatest impact?

Jacob Scott, MD, DPhil: Great question. Not to pitch our own work too much, but if you look back to our 2021.

Dale Shepard, MD, PhD: That's okay.

Jacob Scott, MD, DPhil:Paper in Lancet Oncology, which is... The title is Pan Cancer. It's not truly Pan Cancer because that's every single camp. I think we looked at 11 different types. The signal is there everywhere we look. Obviously, in smaller data sets, it's harder to find statistical differences, but I would say that the study that we're talking about today is our strongest signal in HPV positive oropharynx cancer. The next strongest signal is in triple negative breast cancer, as a matter of fact. And it makes sense because I think that places where our cure rates are lower, we have more opportunity to do better. And so, as an example, hormone positive breast cancers, the signal is less strong for the radiation sensitivity.

We think it's because much of the treatment effect benefit you really get from that hormone manipulation and all of the other treatments are either already maxed out, and therefore, any increase isn't going to change you, or you're just not going to see it until you have massive numbers. So, if I was to predict where this would come into place soonest, it would be HPV positive and negative head and neck cancers. Actually, HPV negative is our next goal. But the data sets just aren't there quite yet. Triple negative breasts, and then stage three lung cancers where the cure rates are not nearly as high as we'd like them, the opportunities exist for us to do better.

And I do think that the first stage of our growth in this field as radiation oncologists will probably be rational dose escalation. As I said earlier, even though we make predictions about places where de-escalation would be appropriate, as physicians and as patients, I think we're going to be more comfortable with rational escalation with the idea of increasing cure rather than the idea of rational de-escalation where we may have the same cure rates and the benefit would be less toxicity.

But you and I practice together all the time. We take care of patients together. And I think that if I was in the position of a patient facing a difficult diagnosis, a lot of times, the question is, "What more can I do?" It's very rarely, not never, but very rarely, "Can I do less and get away with it?" That said, 10 years from now, I really hope we practice in a scenario where we know exactly what each patient needs, or at least if not exactly, we know who needs more, who needs less, who we can treat with less, and get it the same effects. I think that'd be great.

Dale Shepard, MD, PhD: I guess as we look ahead, the 10-year point, what are the gaps? Right now, you're using a particular chip. You're getting developing certain genomics. Do you think there's any other components? Is there some other element that might be thrown into the mix to make this even more effective?

Jacob Scott, MD, DPhil: Yeah. Great question. I mean, I think some of the work that we're doing in our lab is looking at the dynamic changes during radiation therapy. So, this test that we're talking about today is a single genomic test at the beginning of radiation. What my lab focuses on is really how tumors change during treatment. So, the evolution of a tumor during therapy. There are a few different ways you can measure that. One is with imaging. And in radiation oncology, we take CT scans every day on most patients. And so really incorporating the information on how quickly or how slowly a tumor responds. Using that imaging information is something we could add into this.

And then further, I think we're getting better and better at sequencing what we call peripheral fluids. So, for example, can I get circulating cell-free DNA from saliva in head and neck cancer patients? Can I use that as a signal of how it's responding? Could I use peripheral blood? Could I use other easy-to-access points? I mean, saliva is pretty easy. If you could just spit in a tube every week and I know how you're responding, that'd be pretty neat. That's better than a poke, and it's certainly better than a needle biopsy. So, I think what I hope to see and what we're working on in some NIH-funded grants now is to see if we can measure things like DNA and saliva for head and neck cancers because the tumors bathe in saliva, for the most part. Can we use those as surrogate metrics together with imaging for things?

Those are free. You spit in a cup and you're taking an image. There's no needles. There are no knives. Can we use that genomic or radio information to tailor our treatments in addition to how it starts? Because clearly, some tumors will start in a sensitive situation and could evolve to be the opposite, and vice versa.

Dale Shepard, MD, PhD: We're doing some fascinating work. Appreciate your insights. Thanks for being with us.

Jacob Scott, MD, DPhil: Thanks. It's a pleasure to share the information with everybody, and keep your eyes peeled for the paper.

Dale Shepard, MD, PhD: To make a direct online referral to our Taussig Cancer Institute, complete our online cancer patient referral form by visiting clevelandclinic.org/cancerpatientreferrals. You'll receive confirmation once the appointment is scheduled.

This concludes this episode of Cancer Advances. You'll find additional podcast episodes on our website, clevelandclinic.org/canceradvancespodcast. Subscribe to the podcast on iTunes, Google Play, Spotify, SoundCloud, or wherever you listen to podcasts. And don't forget, you can access real time updates from Cleveland Clinic's Cancer Center experts on our Consult QD website, at consultqd.clevelandclinic.org/cancer.