I'm going to talk mainly about some potential complications of radiosurgery, but I don't want you to get the wrong idea. Certainly as Dr. Gertsen has explained very well, radiosurgery really is a powerful tool in our armatarium to treat really an oncologically very, in most cases, a very difficult condition, which is metastatic spine disease. But I thought I would start by disclosing a really odd hobby, which I've developed, which is watching documentaries about airplane disasters. You might think that's a very, and my wife thinks it's a very strange thing for someone who, I mean, most of the faculty here, we travel a lot. I've just been fascinated. I think by the whole investigative process of figuring out what went wrong and learning a lesson, applying that lesson, and it makes air travel for us much, much safer. So actually, instead of being more frightened of air travel, I'm much more reassured that it's very, very safe. And so that's, with that in mind, and I'd like to talk a little bit about the potential complications of radiosurgery. So here's a very interesting incident that happened in Canada in 1983. Air Canada had, was, well, first of all, in 1983, right around this time in Canada, they were switching from the imperial system of gallons, miles, to the metric system of kilometers and kilograms. And in fact, it was very controversial in Canada. People were, you know, up in arms, demonstrations, sign boards, you know, no metric system, etc., etc. So you might imagine it was quite an upheaval in the Canadians, in Canadian society. And Air Canada at this time had taken delivery of its very first metric airplane, the Boeing 767. And so this was their first metric airplane in their entire fleet. And what happened was, about 41,000 feet, on a flight from Toronto to Edmonton, about halfway through, which is in a little east of Winnipeg, Manitoba, the plane completely ran out of fuel. Without fuel, no engines, all their systems shut down. They're running on, their radio was on an emergency battery, but everything shut down. And fortunately, the pilot had a hobby as a glider pilot and knew of and could fly gliders. And they were near an abandoned Air Force base in Gimli, Manitoba, and managed to land without any engine power at all, minimal hydraulics, and all on board were safe. There were really no serious injuries. It was really a spectacular miracle and one of the greatest feats of aviation in Canadian history. And in fact, this airplane is famous, this Boeing 767 is called the Gimli. And for sale, by the way, I just found. You can get it for half a million dollars, a real deal. Anyways, so they go into the investigation to try to figure out what went wrong. Why did this aircraft fly without enough fuel? Well, the problem was they were transitioning from the imperial system of gallons to kilograms. And in their calculations, they actually made a mistake and did a calculation for pounds and ended up only putting in about half of the fuel that they needed for the flight. And so it ran out of fuel halfway across Canada. They took on really 49,000 liters instead of the 20,000 liters that was required because they did have some fuel in the fuel tank. So fortunately, no one had died, but a very small detail led to what potentially could have been an absolute catastrophe. So that's what I want to talk a little bit about this morning, is that it's critical to pay attention to the details. I'm going to give you one case, again in full disclosure, not my case, but it did happen here at this institution. We had a so-called VIP, I'm sure you've all come across these patients, extremely non-compliant, does everything the way he wanted it done. He had metastatic prostate cancer, refused all these things, and my poor colleague didn't know what to do with this guy. Typically we do myelograms to identify the spinal cord, he refused that. And so to in a way placate the patient, this patient was treated without an accurate estimation even of where the spinal cord was. He had a bone-only disease, no epidural disease, they didn't do a myelogram, they didn't do an MRI fusion because he was being difficult, he wanted to be treated right away. So 24 grain, a single fraction was given, assuming that the spinal cord was in the middle of the fecal sac at T4. So here's the plan, looks pretty typical for us, very good coverage, met all our cord constraints with the assumed spinal cord position. But really the spinal cord was not in the middle of the spinal canal as it was estimated. As often is the case in the upper thoracic spine, it tends to drape anteriorly according to the curvature of the spine, and in the upper thoracic spine it usually likes to position itself more anteriorly and not in the middle of the fecal sac as it was originally estimated to be. Sure enough, you can see on follow-up scans there's marrow changes on the MRI scan as you might expect after successful radiosurgery, but this patient also developed cord signal change on the T2-weighted MRI scan. You can see there's edema on the cord. This is one of those things you just don't want to see in a patient in follow-up. So what happened? Well, in fact, if you look at where the cord was estimated to be compared to where it truly was, there was about a four millimeter differential in the cord position. Does four millimeters really make a difference? Well, it certainly can in the realm of radiosurgery where one or two millimeters can make a difference. Here is a dose-volume histogram. A dose-volume histogram is a kind of graphical way to show cumulative radiation doses based on the volume of the tissue or the organ or the target compared to the trace over the dose on the x-axis. So here is in the red what the intended dose of the spinal cord was, and then in green was the actual delivered dose. And so there was a significant difference in the maximum dose received by the spinal cord. We typically constrain our spinal cord doses to less than 14 gray as a maximum dose to any point on the spinal cord contour. But this particular case, the maximum dose received was over 21 gray. We would say that 21 gray to the spinal cord in a single fraction, you're in tiger country there. Some interesting data from pigs. I think this is really my favorite data set on spinal cord tolerance. This is, I think, a very elegant set of experiments done on mini pigs by Paul Medin at UT Southwestern. These mini pigs apparently have the most anatomically similar spinal cord and spine anatomy to the human. And so what he did is he took a series of pigs, did a dose escalation experience using stereotactic radiosurgery techniques, which means it was partial cord irradiation to high doses. It wasn't full cord. It was just partial cord the way it would be for radiosurgery, and then did a dose escalation exercise from 14 gray to, I think it went to 24 gray in a single fraction. And the red line represents the probability of histopathologic damage. And so what he did is that he took these spinal cords, sacrificed the animals, and did a histopathologic analysis to look at the amount of damage that was inflicted from the radiation. The confidence intervals are relatively wide here because he used only three or four pigs per dose level. But what comes out, I think, is a very important lesson for us as we pay attention to the details. And that is that the complication probability curve is very, very steep. It starts off really less than 17 gray, zero probability of spinal cord injury. But when you get above 21 gray, you're almost absolutely certain to have spinal cord injury. So over a range of about four gray, you go from zero to 100% complication probability. What does that mean to us as clinicians? It means we have to be extremely careful because once you enter into tiger country, you're all in. A few millimeters can make a difference in dose of 10% to 20% to the spinal cord. And so a setup error or a mistake in calculation, or there are a number of things that can happen because this is very complex, a minor mistake can end up resulting in a catastrophic result. Again, I don't want you to get the wrong idea because we know that stereotactic radiosurgery is a very powerful tool to benefit our patients. But just like any other tool, you have to be careful. And we are careful. And because of that, the complication rates from spinal radiosurgery are extremely low, particularly for spinal cord, certainly less than 1%, much less than 1%. But it underscores the importance of paying attention to the details. How about even in the re-irradiated setting? This is equally interesting. For spinal cord tolerance, in this experiment, Paul took a number of pigs, gave them all 30 grain, 10 fractions, and then a year later, repeated dose escalation exercise. And he found that the spinal cord tolerance in pigs who had previously had conventionally fractionated radiation to the whole cord this time, it was a standard technique that he used to give 30 grain, 10 fractions. A year later, it didn't do radiosurgery, absolutely no difference between spinal cord tolerance of pigs who had never been irradiated and pigs who had had radiation with conventional fractionation a year prior. The curves are almost superimposable. But the point here is that the dose where you would have a complication of either a 10% risk or a 90% risk is over an extremely small, short interval. Here are some spinal cord tolerances that various institutions that have a very high volume of radiosurgery would use. I think it really comes around to about the same thing. We're all focused in on about the same level in terms of what we will accept to the spinal cord. But in full disclosure, again, this is not based on any science, really. But we have found that with these kinds of cord constraints, if you pay attention to the details, that it's extremely safe and that you can achieve much higher levels of tumor control if you follow at least these types of guidelines, the rate of cord complication is extremely low. But if in order to achieve this, you have to be able to work well as a team, with as radiation oncologists or neurosurgeons, also with medical physicists, and every link in the chain has to be equally robust. We've actually found that our biggest complications are actually esophageal. And here's a case of a 45-year-old patient, oligometastatic renal cell carcinoma, got 24 gray in a single fraction, and received 15 gray to 2 cc's of the contoured esophagus. And this patient has this type of a dose volume histogram. And the green contour here, or the green line, represents the dose received of volumes of the esophagus. And you can see that, especially in the thoracic spine, it's very hard to avoid the esophagus if your intention is to treat the entire vertebral body, which we would recommend as your clinical target volume. And in these days, when we treated this patient, we weren't fully aware of the potential complications that radiation can cause to the esophagus. We were really focused on sparing the spinal cord. This patient developed esophageal pain, eventually was found to have a non-bleeding ulcer at the level of treatment, and an outside gastroenterologist was worried about it because he wasn't familiar with this type of injury, so he did a forceps biopsy. Worsening pain, increased ulceration and infection, this just didn't heal. Again, the gastroenterologist this time did a dilation and did another forceps biopsy, and then by 6 1⁄2 months, he developed a tracheal esophageal fistula. You can see there's an attempt to put a stent in there on this CT image here, underwent multiple repairs and stent procedures, and this patient really could not have systemic therapy in this condition. While his oncologists are waiting for his esophagus to heal so he could get chemotherapy, his disease progressed, and he expired from disease progression. In this case, it's not very comforting to tell him you died of a metastatic disease, but at least your thoracic spine disease was okay. It's a hard sell that way, and I can't help but think that this poor patient's esophageal problems really led to early demise. And we've learned a lot of lessons about esophageal complications with radiosurgery. This is a paper that we published a couple of years ago now where we had seven patients who had grade four or higher esophageal complications, well, not all of them, but most of them at the 24 gray level. What we found was that there were no serious adverse complications to the esophagus without either anthracycline-based chemotherapy and or iatrogenic interventions, particularly biopsies. These kinds of things should be stopped. There's no purpose for it, but you may have outside gastroenterologists who are just not familiar with the effects of radiation on the esophagus. And I think just about anybody who has radiosurgery, if you did do an endoscopy, you're going to see mucosal changes. And when they get multiple biopsies, especially in combination with anthracycline-based chemotherapy, those patients are really at risk. And I tell all my patients, do not let anybody touch you unless your doctor talks to me first. We feel very strongly about this. And it was a hard lesson learned. This is another interesting aspect of esophageal dose tolerance. And we created a complication atlas based on all our patients who had radiation really within the levels of the spine where there is esophagus. And we took all their dose-volume histograms and piled them on top of each other and created a map. And this is the probability of a grade three or higher esophageal complication based upon the volume of the esophagus on the y-axis and the dose on the x-axis. And so if you have part of your dose-volume histogram that goes into the red, then according to our database, you have nearly 100% likelihood of developing at least a grade three esophageal complication. If your dose-volume histogram falls within the darker shades, then that risk is significantly lower. And if your dose-volume histogram would fall within the green area, so this is 20 gray, by the way, right here, then you have up to a 50% probability of an esophageal complication. So this is just a culmination of our experience mapped against a specific complication. So we've also found this type of tool to be very helpful as we assess our plans and ask ourselves, what is the risk of a serious complication if I accept this plan for treating this patient? Overall, we found by changing our esophageal normal tissue constraints to 14 gray to two and a half cc's that that seems to be a cutoff. And we made this change about two, I guess about two or three years ago now, and we have not seen a serious esophageal complication since we've made this change. Now, you might think that going from our previous limit was 15 gray to two cc's esophagus and going to 14 gray to two and a half cc's, could that really make that much of a difference? Well, I'm here to tell you, clinically, we have not had a case of grade three or higher esophageal complications since we made this dosimetric change. And again, I believe it's paying attention to the details that makes a difference. Compression fracture, I'll probably stop after I talk about this. This is actually probably our most common post-treatment effect that we see. Here's a case of a 63-year-old patient with non-cell cell lung cancer who had radiosurgery to T5, and going from October in 2006 to March of 2007, this patient developed acute onset back pain, and here you can see that in about, within six months of radiosurgery, this patient has a compression fracture that's developed at the site of treatment. I'm just gonna quickly go over a paper. It was really the first paper that addressed this observation in radiosurgery in the spine. A JCO paper from 2009, looking at 71 patients, or 71 lesions, treated from 18 gray to 24 gray in a single fraction, and we really follow these patients as if they were on protocol. We image them every three or four, well, I'd say now, every three to six months, and essentially got our best neuroradiologists, locked them in a room with three spine surgeons, and gave them lots of pizza. Actually, I don't know, was there any pizza there, Mark? No pizza, okay. They were faster. Yeah. Yeah. And looking at, really, as a primary outcome, radiographic progression, or either of a new fracture or progression of an existing fracture. And then we looked at a bunch of other clinical factors, such as the ASIA score, pain scores, and narcotic use, and what we found was that there was gross radiographic change in almost 40% of our patients who had radiosurgery, and the CT appearance was a significant predictor for vertebral body collapse, where lytic lesions were almost seven times more likely to fracture than either mixed sclerotic lytic or sclerotic lesions, and the percent vertebral body involvement was a significant risk as well. And the more vertebral body that was involved with tumor, the higher the likelihood that your patient was going to undergo a compression fracture, either progression of a previous fracture or a new radiographic finding. So I want to stress that these patients are, we really reported on radiographic changes. Not all these patients are going to become symptomatic. That T5 lesion, I think, is a little bit of an outlier, because most patients who have fractures in the thoracic spine, particularly away from junctional areas, remain minimally symptomatic or asymptomatic, and there's really nothing that needs to be done. It has no impact on their quality of life. We just see it radiographically. However, some patients do become symptomatic. So here is stratified by CT appearance, lytic versus non-lytic. You can see in the dark line, the non-lytic, they tend to fracture later and not as frequently, but follow these patients long enough, and the majority of them will demonstrate radiographic changes over time. So we're just updating our patients who have had at least five years follow-up following radiosurgery, and I think this has been presented as an abstract, but with patients who have had at least five years of follow-up, again, our results are similar to what we published over five or six years ago now, which is that of almost 40% radiographic fracture rate, about 12% overall required some kind of intervention. So about 12% of patients who had radiosurgery and have at least five years of follow-up have required some kind of intervention like a kyphoplasty or vertebroplasty, and that represents about 30% of our radiographic fractures require an intervention. Others have looked at this. This is from MD Anderson, looking at patients who either had 18 gray in a single fraction, nine gray times three, or six gray times five. They found about a 20% risk of compression fracture after SVRT, and in their multivariate analysis, the only significant factors were age greater than 55, a preexisting fracture, or preexisting pain. All other factors, including dose, that did fall out in multivariate analysis, and certainly in our analysis, dose was not a predictive factor, though I know there's a paper from Toronto where they do feel like the dose of radiation can make a difference. So we always talk about an initial presentation with patients, the probability of a compression fracture after radiosurgery as a real potential complication, although we believe that ultimately, significant proportion of those patients will never require any kind of intervention for the fracture. Okay, I've got some other stuff, but I think I'm just gonna stop there. I think Dr. Bielski will talk a little bit about the relationship of dose and outcome. So I'm gonna stop with that, and take any questions if there are any.

radiosurgery

complications

spinal cord estimation

esophageal complications

compression fractures

research