Challenges and Dilemmas in Adaptive Radiation Treatment Planning
This on-demand video offers a quick overview of our program and the real-world impact of adaptive planning. We’re now able to treat patients who couldn’t have been treated before—opening the door to new possibilities. And with deep experience in this approach, we’re helping lead the way forward.
Hello, I'm Eric Horwitz and I'm the chair of radiation oncology at Fox Chase Cancer Center and the Lewis Katz School of Medicine at Temple University. Today, I'm gonna be talking about our adaptive radiation therapy experience at Fox Chase and some of the interesting cases and experiences we've learned over the last several years. This is a picture of our first treatment um at the end of February of 2023. So the goal, some of the goals of my talk today is to one, identify the different reasons to use adaptive radiation therapy and who are the candidates for this treatment. I want to discuss situations where adaptive radiation therapy is the only way to safely treat a patient and discuss some clinical examples where adaptive radiation therapy was needed. So I would like to start off with what's the fundamental goal of modern radiation therapy. It's basically to deliver a high dose of radiation to the cancer, the target, and minimize as much as possible the radiation dose to the surrounding normal tissue. This has been a goal of radiation oncologists and radiation therapy for generations. So, I'd like to say, what are some of the primary reasons to use adaptive radiation therapy? So, in our experience, the first one is the adaptive treatment plan will provide superior target coverage compared to a non-adaptive treatment plan. And, or the adaptive treatment plan will provide superior normal tissue radiation dose avoidance in one or more of the adjacent organs at risk compared to a non-adaptive treatment plan. It can be one, the other, or both, but it needs to be one of these. So who's a candidate for adaptive radiation therapy? So in our experience of Fox Chase, it's. It's a lot of prostate patients, certainly are prostate SBRT patients, some of our bladder patients, liver patients, pancreas and biliary tree patients, um, patients with oligometastatic disease, both in the lymph nodes and bones, some lung and mediastinum cases, cervix patients, kidney patients, head and neck patients, and some IMRT patients, um, the cervix patients, the prostate bed patients, basically, anybody who's a candidate for SBRT is, is probably a candidate for adaptive radiation therapy. But what we've learned over time is that there's another group of people that are also, you know, there, there are reasons then to treat these people with adaptive radiation therapy. And some of these reasons are, we couldn't safely treat them if we didn't have adaptive radiation therapy. And the other group is actually rear radiation patients, patients who have already had radiation, and the adaptive treatment plan really provides crucial normal tissue dose avoidance, um. Compared to a non-adaptive treatment plan where we just wouldn't be able to treat if we didn't have the adaptive technology. So before I get started, I just kind of want to talk about some of the basic nomenclature, since I'll probably use it in different ways. So, in a CT adaptive world, the reference plan is the original treatment plan that's generated from the simulation. The scheduled plan is this original treatment plan that's recalculated based on On the combing CT of the day, um, which is shown side by side with an adaptive plan, and the adaptive plan is the new plan generated on the day of treatment from the new data set with modified and approved contours that reflect the changes in the the volume, the position, the um, the, the location. And finally, OARs organs at risk. It's the adjacent normal things next to whatever the target is. So, as I kind of already said, so this is, this is the group of patients that we routinely treat with adaptive radiation therapy. So it's prostate SBRT, liver, pancreas, patients, breast APBI, oligometastatic lymph nodes, some oligometastatic bone meds, kidney SBRT cervix IMRT, prostate bed IMRT, and there's some others, but this is kind of the main group. But what about the patients who are not routine? What can we do for them? So here's where I have now some a bunch of examples and over the last several years that we've, what we've learned with the adaptive technology to be able to treat patients we normally wouldn't have been able to treat. So the first problem was how do you treat a patient who has two sites of abdominal oligometastatic disease on the same side as their only kidney. Hm. So this patient was a 68-year-old who had metastatic prostate cancer and it had um positive lymph nodes near L1 and L4. The plan consisted of two ISA centers. We were planning on giving 800 °C x 5 for a total dose of 4000 °C, but they only had one kidney and of course it was on the side where the disease was. So this is an example of the original treatment plan. And this is the first fraction of the this is the, this is the case for all 5 fractions. So here you have the, uh, the new treatment plan and you have the side by side comparison of the adaptive plan and the scheduled plan. And in this plan, there is significant improvement in the target volume coverage of both the, the egocy disease at L1 and at L4. But at the same time, there is dose reduction to the normal kidney and the spinal canal. So we actually got much better coverage to the targets, and we, we spared more kidney. Um, and this was ended up being the case for all 5 fractions. So the second example now is, how do you treat patients with a clinically significant mass that's between both small bowel and rectum. So this is a case for the perirectal mass. So here we have another prostate patient, um, and bladder patient. So the patient had a history of pathologic stage 3 prostate cancer, poster robot assisted laparoscopic prostatectomy, and then they had a stage 2 transitional cell carcinoma of the bladder, and they also then had a cystectomy. So now they end up with a locally recurrent prostate bed mass. And this mass, because of both of their two surgeries, is literally between the small bowel and the rectum. So our plan, what we wanted to do was to give 700 °C times 5 for a total of 3500 °C. The patient's pre-treatment PSA was 6.1. They've been on hormones intermittently, but they really weren't tolerating them, and they really, really wanted to be off the hormones because it was causing a dramatic effect on quality of life. So how do we safely treat this mass but protect their small bowel and the rectum? So this is. This is their adaptive plan and one of the things that you that we do differently with adaptive radiation therapy compared to non-adaptive radiation therapy is we set the priority before we do the planning, and you have to tell the physicists this. But historically, the target always had priority. You wanted the cancer to get as much radiation as possible. But in this situation, we actually wanted to protect the small bowel and the rectum, so they actually have priority over the target. Um, so then we generated a plan, and now we do the adaptive plan each day. This I took an example of fraction 3, and here you've got target coverage is pretty much the same compared to the adaptive plan and the and the the pre-plan. However, um, on the adaptive plan, there is a significant reduction in dose to the rectum and you can see here in this column right here. Um, and again, all the fractions showed this where it was between different points and different volumes between 50 and 100 °C less per fraction. So dramatic reduction. Um, also, we saw a significant dose reduction to the small bowel as well. So this is ideal. Less dose to the rectum, less dose of the small small bowel without compromise to the target. Um, we treated this patient, um, um, almost a year and a half ago now, um, and this is the patient's PSA, um. Actually at 20 months, which is significant. He's not on hormones, so it really worked. He has, he's had no change in his bowels, no real significant side effects. So this is what we wanted to do. So another example. So how do you treat a patient who needs palliation, but is also paralyzed and requires significant support just to get out of their house and into the hospital and the radiation department. So this is a patient who had cholangiocarcinoma, had paralysis, was in a group home. Um, and so 78 year old man, he had a history of cerebral palsy. He had some dementia. He was Hoyer lift dependent and lived in a facility. We just wanted to give a single fraction, but even giving a single fraction was a significant was significantly burdensome. So in this situation, what we did, it was gonna be too difficult to get them to come in even for the simulation and then coming back another day for the treatment. So we took their diagnostic CT and did the plan off of the diagnostic CT because we needed something to be able to compare to the adaptive plan. We used to identify for surface tracking because there was a lot of movement. Um, and this is the treatment. Here's the example, um, the single fraction, and we had excellent target coverage. Um, we, we, we were hitting where this big mass was. Um, and we had good dose of avoidance to the esophagus, the heart, the spinal canal. The key was is that we didn't do any pre-planning or we didn't do any different preplanning. We preplanned off the diagnostic scan. We brought them in. We did everything on the machine. We got him treated. He was just one visit to the department, and we got excellent coverage of the target while we missed the adjacent critical structures, and we were confident that we missed the adjacent critical structures. Um, also had good dose of avoidance to the stomach and the liver. So another example, how do you treat a patient with oligomedesac disease that surrounds the critical organ at risk? So this is another kidney patient. And here we have a a seven year old person who had metastatic hepatocellular carcinoma to the left adrenal gland, and you can see here that um there's a little bit of kidney here and it's surrounded by tumor all above and below it. So again, we wanted to give 4000 °C in 10 fractions because we needed to to spare the kidney as much as possible. And of course, they only had one kidney. We seem to have a lot of these patients. So the planning priority again was the organs at risk, the normal kidney tissue, the duodenum, the bowel. And here again, we set the uh the priorities in the in the pre-planning to um to limit small bowel and duodenum coverage. And here's, um, here's one of the fractions, and this is the 9th fraction. We adapted all 9 fractions, and here you've got good target coverage again. It's, it's pretty similar. There's not really, it's not that different, but, um, we were really able to, we had lower dose to all of the organs at risk, the small bowel, the duodenum, and the kidney. So again, it's ideal. We, we hit the target, we missed the, the, uh, the significant, the crucial, uh, adjacent organs at risk. Another example, how do you treat a patient who has clinically, a clinically significant hiatal hernia that moves into the target volume during treatment? And I think this is an example now where these things probably are happening all the time and we're not even necessarily aware of it or if we're aware of it, there's really not much we can do about it. So we just kind of go ahead and treat or we don't treat, but we can't, can't adapt, we can't adjust. So this is a patient who had lung cancer and a really significant hiatal hernia. Um, and here on the imaging, you can see how big this hernia was and it was totally up in the chest. And it was an 86 year old person who had a stage 2 non-small cell lung cancer. Um, the hiatal hernia wasn't amenable to repair. There wasn't any way to get it to move out of the way, even in different positions. And the plan though was to give IMRT to the left lower lobe and the hilum, 6000 sun gray, 30 fractions, but this hernia was gonna be there all the time. So what do we do? So we made the hernia be an an an organ to avoid. And here's an example. Um, this is the happens to be this is the 29th fraction of the 30 fractions. But again, you've got improved target coverage with the adaptive plan, so it is. Better, um, but crucially improved dose avoidance to the stomach with the adaptive plan, and we use the adaptive plan for all 30 fractions. And you can see, and the other thing to note this I think when we talk about an IMRT plan versus an SBRT plan, sometimes the, the dose improvement isn't as dramatic for the IMRT plan compared to some big SBRT fraction. But if you're gonna Multiply it over 30 fractions or 10 fractions. That difference in dose is going to add up over time and it's really going to result in a dramatic reduction over the whole dramatic reduction in dose over the whole course of treatment. So again, for this patient, we hit the target, we got exactly the dose we wanted to, and we really were able to spare the stomach by hundreds and hundreds of centigrade over the course of treatment. So now this is this is another thing I think we actually run into more than we think, and a lot of the time, we'll try to do the best we can. But what happens is if your patient can't be supine or prone. There are plenty of patients like this from many different cancers, and You know, we do the best we can. But So in this case, we had a man who had a metastatic renal cell carcinoma with a peritoneal nodule on the right side, but due to multiple medical comorbidities, the patient wasn't a candidate for systemic therapy. We needed to do radiation, but he couldn't lie s supine or prone. He could lie on the side. So the plan was to give 4200 °C in three fractions, but how do we make this be reproducible when they're in a completely different anatomic position? So we adapted, um, and this is an example from fraction 2, and what we saw in fraction 2 was, um, That we got better, we got absolutely better coverage with the adaptive plan to this to the target right down here. You can see it actually in all the different um categories, got more dose, um. And um we were able to spare the kidney and the duodenum, and we also did it with confidence. We could see for each of the fractions the beam was going where it was supposed to be going, and it was missing where it was supposed to be going. So in this situation, we had better target coverage and we had better dose avoidance, and it was reproducible because we just replanted each time and it was only 3, it was 3 fractions. And this is just an example again from fraction 3, but we saw for all three fractions. So now it's another situation where I think we see it. It's probably more common than we appreciate. And again, we usually do the best we can, but now we don't have to guess, we know exactly what's happening. So in this situation, what do you do if a patient has acites and If you if you remove the ascites, but you do that in between either the simulation and the treatment or between multiple fractions, it completely changes the anatomy dramatically like in it, so, you know, how do you take into account this change in anatomy and this change in position to the target volume? So in this case, it was a patient who also had a had a cellular carcinoma, the oligometacytic disease in L5, and the plan was to give 2700 °C in three fractions, and they had massive ascites. So between the simulation and the first fraction, they took off 13 L of fluid, um, and between fractions 2 and 3, they took off another 6.5 L. Um. And it completely changed the location of the target and all of the all the normal structures. And it was in fact this is something you know we don't see that often, but it was so dramatic that the plan that the system couldn't take the scheduled plan, which is the pre-plan superimposed on today's anatomy. It just didn't even come close to matching, so it couldn't even generate a scheduled plan. So we had to use the adaptive plan because the adaptive plan was what was in front of us. We recontoured everything, recontoured the normal structures, we recontoured the target, and we did a plan. So we got a great plan. Um, but it couldn't, it was such a dramatic change that it couldn't even do a comparison. So if we hadn't had the ability to do this, we could not have treated this patient because we just would have been treated in a completely wrong area. So here's an example from fraction 2 where it did generate because there was less fluid. It did generate a scheduled plan, but um but so the adaptive plan was much, much better. Um, and again, getting a better target coverage. And at the same time, better, much better dose of avoidance in this situation, it was to the sacral plexus, um, but again, we're hitting the target and we're missing the normal stuff. Here's the bell. So, another example, um, how do you treat a patient who's already received a full radiation dose of the targeted volume, but they have no other treatment options for their recurrent disease. So we're talking about rear radiation, and this is something in the past we didn't really do as much rear radiation, but what we have absolutely learned over the last bunch of decades is, in fact, we, you know, the body that can tolerate more radiation, but we always have to be precise. But with real radiation, we have to be even more precise because, you know, the, the adjacent organs at risk have absolutely already been radiated. So this is an example where this is a patient who had locally recurrent high grade liposarcoma. They were treated in 2023 to the left inguinal area and pelvis, receiving 3800 °C 19 fractions. They got a boost of 2200 °C 11 fractions. And the plan was now to treat this clinically significant left groin recurrence for another 4000 °C in 5 fractions, but it was gonna overlap with the previous treatment. So here what we did was we identified the target and we made the all the all the adjacent organs at risk, the bowel. It would have been a a dose avoidance area, but we made it be that it was the primary uh dose avoidance, that the that the primary priority. We needed to miss the bowel that had already gotten some radiation. Um, and this is an example of fraction too, where the target coverage, coverage is better, not dramatically better, but it's better. Um, and now at the same time though, better dose avoidance to the bowel, and with these were all within acceptable limits, and you can see this is just a single fraction, but it's less than 100 cent. The bowel got less, less than 100 °C with the adaptive plan compared to the non-adaptive plan. So again, clinically significant. This is something certainly in my world where I'm frequently treating all these prostate patients, and we're always talking about that we want patients to have a full bladder, do the best they can, but this is an age-old struggle for our prostate patients and our prostate bed patients. Um, you know, having a full Uh, a full bladder makes the treatment more producible, reproducible, but what if you can't have a full bladder, and this happens every single day in our. Clinic. So this is a prostate bed patient who basically had no bladder filling. They just, they were incontinent. They could not short of us clamping their bladder, you know, um, you know, the urethra, filling the bladder, doing all this stuff. They could just never have a full bladder. But if you're adapting, it doesn't matter because we're just gonna, we're gonna plan what's in front of us. So here's a patient who had pathologic stage 3 prostate cancer. Um, their poster in PSA was 0.4 and rising. They had a high Gleason score. They'd had PSMA PETs, so we know it was, it was positive in the bed. We know we wanted to treat the bed, um, but they could not fill their bladder. Um, and here, this is, this is imaging where, so this is the best they could do, and it's, it's just a tiny little bladder. Um, they had actually been told in an outside hospital that they just Couldn't get treated because they had no bladder filling. Um, so the recommendation was, um, hormones only, no radiation, but this is a relatively young, healthy person. We expected him to be around for a long time, and hormones alone in this situation is not really adequate. So the plan was to do moderately hyperfractionated adaptive IMRT. We're going to give 6250 °C 25 fractions to the prostate bed. Um, and here we, we adapted all 25 fractions. The target coverage was in all of these fractions was better or the same. The rectal dose was always better, and the bladder dose avoidance was sometimes better. So, um, we always made sure we hit the prostate bed. We always gave less dose to the rectum and a significant portion of this treatment, we gave less dose to the bladder. Um, and we're also confident that we're hitting the target where we wanted to be hitting the target. Um, so again, this to me is an ideal situation. And here you can see an example target coverage is better, not dramatically better, but it's better. Um, but the rectal dose avoidance was always better. And I think this is a function again of when the bladder fills, it, it, it can move the rectum, it moves, and it moves where the prostate bed is. So again, we knew where it was, we knew, you know, we knew where to hit the target. One of the final examples is you have a clinically significant recurrence and you need to protect the brachial plexus, because you can really have catastrophic damage if you overdose the brachial plexus. So this example is 49 year old person who has recurrent metastatic lung cancer originally treated in 2023, um. To the right neck, 6, 6000 sun grain and 15 fractions. They now have a recurrence in the mediastinum and lungs. So the plan is to give them another 4000 sun grain and 15 fractions. The priority, the influence structure, which is the normal things that was given priority with the brachial plexus. We, we were under the limit from the original course of treatment. We could not overdose the brachial plexus. This is one of our first situations where we realized that we wanted the the normal structure was gonna get priority over the the target. Um, this is an example of fraction 3 of the treatment. And here, it's kind of hard to see, but you've got all the targets and you've got all the normal structures, the spinal cord, the brachial plexus, the larynx, the great vessels, skin, all these different things. And we had improvement of the target coverage. We were hitting what we wanted to hit, um, with the adaptive plan, um, and. Again, a drama a significant dose of avoidance to the brachial plexus. And again, this is one fraction of 15, and you can see those fractions will add up to a dramatic dose reduction to the brachial plexus. We could not have done this if we didn't adapt. And I think another 11 of the final examples now is again with kidney, and there's a lot of cases with kidney where how do you treat a kidney primary but spare enough normal kidney um to prevent prevent or delay dialysis and systemic therapy. So here's an example of the patient who has a stage one renal cell carcinoma with an atrophic left kidney. Um, renal cell is, is, you know, 4 by 3 by 3 centimeters, and the treatment goal is to spare as much of the normal kidney as possible and to delay the initiation of systemic therapy. So, and they have an atrophic, um, right, uh, atrophic left kidney. So the plan is to get 4200 °C in three fractions to the right kidney mass, um, but spare the the normal kidney, you know, uh, uh, the normal tissue in that kidney. So here's fraction one. Again, we went with the adaptive plan. So a slight improvement in target coverage. We're happy. It's, you know, it's a little better, not much, but it's, it's better, but a significant improvement in um the dose avoidance of the normal functioning kidney. And again, you can see it's just a, a significant amount of volume of the kidney is, is getting less volume is getting radiated. Um, also, in addition to sparing the kidney, we actually were able to spare the stomach and the adjacent small bowel. So it's a win across the board on sparing the adjacent normal tissue at the same time, we're hitting the target. So this is some of the lessons that we've learned from, from our adaptive radiation program as it's matured, and as we've moved beyond what we would consider our traditional adaptive cases into our now more nontraditional adaptive cases and our special cases. So, one, adaptive radiation capabilities has enabled us to safely treat patients that we would not have been able to treat in the past, um, either for dose of dose avoidance to critical adjacent normal structures, or just to be able to re-radiate a target that we just wouldn't have been able to do before. Um, an ideal adaptive plan would be, would be an improved target volume coverage and significant dose dose reduction to normal surrounding tissue, but we don't always get that ideal adaptive plan. Sometimes the target volume coverage is improved, sometimes the normal organs are dose reduced. Sometimes the target volume coverage isn't improved, but it's not bad, but the normal organs are significantly improved, and that's a reason to do the adaptive plan. Um, all of these things require clinical decision making to be to select either the adaptive or the scheduled plan, and these are the things you learn as your program matures and you do in more and more adaptive cases. Clearly the physicians and the physicists must collaborate to set the clinical goals. The physicist or whoever is doing the the planning of the dosimettris, they can't read our minds as the physicians. They need to know what's the priority, the target or the normal stuff, and you need to prioritize and select what's the most important goal. So I want to thank you for your time. Here's a picture of our original core group, although it's included in most of the department now, our team. Here's the day that the uh ethos got delivered. And again, thank you very much for your time, um. Have a good day.
