So, my talk is really centered about evidence-based radiosurgery, so I'm going to talk probably a little bit more rather than about some specific data. I'll give a little bit of specific data, but really talk about the methodology a little bit. And I think it's been something that's been really better explored, and I think in a more organized way in the neurosurgical society in terms of developing recommendations and guidelines using this methodology. No disclosures relevant to this talk, and in terms of the process, essentially, for any evidence-based method, you start out with a defined question, and then what you want to do is you want to pull the literature, and you want to get as much of the literature and agree to kind of what your predefined notions are. So, you may eliminate case reports. If you don't have survival outcome, you may want to eliminate that if that's one of your, you know, study questions. If you don't have enough prognostic information in it, then you may want to eliminate those studies. And you basically get a group of experts in a multidisciplinary field to agree to do this, and the AANS has set up several very, you know, nice, relatively large committees that have worked on this and ended up publishing a number of papers on that. When you do this, then each paper is assigned a class, and that's basically based on the quality, and then the outcome of what those are based on class are put into evidentiary tables, and they're reviewed by the committee to agree what are the real evidence-based conclusions you can have based on class one, two, and three data. And I'll go through what those are in a minute. So, again, thorough review of the literature. Try to be as comprehensive as possible. You really want to weight it on scientific validity and directly link it to the evidence, okay? You're really trying to be a stickler about this. We all, okay, even myself, am actually biased, and I have some very, you know, specific biases whenever I look at a given amount of data, and so this methodology really tries to take that out. Expert judgment is really given a secondary seat to what's there. However, you have to recognize that most things that we do in medicine, frankly, are not based on class one data, and a purist might argue that, you know, that's bad, but I think that that's reality. I do think these help in terms of helping define what areas that maybe we should study a little bit better. I think sometimes, however, it, as I'm going to show a little bit later, sometimes it can be an excuse not to move the field forward or to use knowledge that you actually have, and I don't think you should use evidence-based terminologies to kind of avoid progress as well, which I think sometimes is the case. So with that said, level one, very high degree of clinical certainty. Usually these are randomized control trials looking to answer a specific question, and then they actually answer that question with a relatively large, you know, RTOG type study. It can be an institutional study, but you really want to have it randomized. These are obviously multi-institutional is better. Level two is really the second highest level of evidence, and that being a moderate degree of clinical certainty. A lot of times there'll be some more robust comparative studies. There are a number of studies that were done early on with University of Wisconsin, UCSF, University of Florida, all participated where we took a given group of patients, tried to look at a case control study for this group of patients, and then compile all the data and look at a specific outcome, and that would be a good example of a type of class two type evidence supporting something. So stronger than three, but not quite as good as a true randomized control trial. Usually it's looked at a little bit in terms of retrospective. Level three is really institutional studies, and the reality about institutional studies is we have a lot of things we do where there are multiple institutional studies. There are, you know, when you get into multiple, multiple institutional studies all showing the same thing, I think you reach a conclusion, you know, that's probably true. And so I think we shouldn't just say, well, that's level three evidence, therefore it's not worth anything, and I think that's another thing we want to avoid when you think about evidence-based guidelines and conclusions. In terms of what we're going to talk about today, really we're going to look at it in the realm of treatment. We want to see which of these treatments can actually be better, surgery, you know, radiation therapy, or I'm going to use the example really of whole brain radiotherapy, and then chemotherapy and radiosurgery. And so you're going to basically ask the question in terms of what's in the literature about some of these methods. Again, we went to level 1, 2, 3, class 1, 2, 3 basically followed the same thing in terms of therapeutic effectiveness, it's class 1 evidence, 2, 3, they follow each other. These are some of the examples of some of the papers that have come out from the AANS and I think are just examples of the kind of thing that can be done. This is my favorite slide in this, and it's because I really think that, again, as I highlighted earlier, a lot of people try and use evidence-based guidelines as kind of, you know, this kind of argument that essentially there's never been a randomized study to see whether or not you need a parachute if you jump out of a plane. And the reality is that there hasn't been, nor is there likely to be one, and this kind of was a satirical article that went through and kind of, you know, compared this and looked at all the rationale from the statistical standpoint of why this should actually be done and they suggested that most of the protagonists of evidence-based medicine should be organized to participate in the double-blind trial of this. So it's, again, a little bit tongue-in-cheek, but there are some things that we really, really do know, and I'm actually going to maybe even use this because I'm going to argue that, you know, some of the trials that, you know, Eric, you know, very nicely presented where we're looking, and we're going to unending, you know, and we're spending a lot of money doing neurocognitive studies for patients with whole-brain radiotherapy to see if they have cognitive deficits, you know, and I don't know, I mean, I think that we have some pretty good evidence that if we do whole-brain radiotherapy on patients with whole-brain therapy, I mean, there's long literature of this, and I, you know, there's just, I don't think there's a lot of doubt, and so we, I think we can, you know, question. Sometimes we go to the extreme in terms of evidence-based proof of what we're doing, and when you're dealing with some of these questions, I at least would raise that. So if you look at, let's consider brain metastases, whole-brain radiotherapy, increased survival about from one month to three months. This was an institutional study control back in 1954. It was followed up by really large, randomized, maybe class one type data looking at a whole group of different fractionation screen, basically showed the same results, showed that the fractionation schemes weren't a whole lot different from one another, but, you know, certainly showed that the radiation therapy and whole-brain radiotherapy itself was effective compared to nothing. So it confirmed efficacy. It really, 30 in 10 fractions in two weeks became a de facto standard, although I think you could have, you know, argued a little bit towards one or the other, and you got clear improvement. You got, you know, better long-term survival. I think that there's very good class one data for doing surgery. If you look at Patchell's study, he did surgery plus whole-brain versus the whole-brain alone and basically found that surgical resection with solitary metastases led to a significantly better outcome in survival. This is a good example of why that is true. And then did the other study where they eliminated the whole-brain and showed that the whole-brain was needed if you were going to increase control, so a good example of a really class one data. Well, out of that concept, you know, there's a concept that came through Patchell's study, which was that if you could gain local control, you could prolong survival and actually do better in patients with brain metastases, and we started developing a pretty significant literature of class three type data, and you had institutional studies coming from, you know, Boston and Wisconsin group, University of Florida group. We had some initial papers just looking at radiosurgery as a way to gain local control and didn't have a lot of randomized data. That led to multiple randomized studies where you had patients who essentially were treated with metastases, you had Kanzioka treated two to four metastases, whole-brain versus whole-brain plus radiosurgery, and basically the local control, much better with radiosurgery, not unexpected. The RTOG study, as previously cited, did the same thing, whole-brain radiotherapy versus whole-brain plus radiosurgery, again, much better, better survival in those with solitary lesions, better control in all the lesions, but not clearly a survival advantage in some of the ones if they had more metastases. If you look at radiosurgery alone, and this is probably, you know, where I would come down, if you look at radiosurgery alone, you have, you know, one to four metastases treated with radiosurgery alone versus with whole-brain, and there's absolutely no survival with whole-brain. There are more recurrences in the radiosurgery group, but if you're screening and you at least allow treatment, these patients do just as well, and there are about eight or ten institutional studies. So I would actually argue that for one to three metastases that radiosurgery alone is the de facto standard, and I guess I wouldn't agree, but again, I recognize there's some controversy, and if you go with a lot of radiation oncologists, some people are still saying, but I think there's tons of data, and, you know, I always give the example, if my head had two metastases, I get to, you know, I get to lose all my hair, which, you know, maybe I lost anyway, but, you know, and I'm going to lose hearing, and I'm going to have these other stuff. Do I need to have a randomized study to say whether or not I would get that? I've treated a lot of patients for a long time, and I wouldn't take it, I mean, personally, if I had, you know, one or two metastases with a solid tumor, and I think most people doing radiosurgery would say that same thing if you go through it. Now I think the question is when it gets to the four and ten metastases, when do you really have to stop it? We don't know those limits, and so our limits have increased as we've gone through, and so it's not uncommon to treat four or five, you know, maybe we can do six to ten, but, you know, where is that limit? And I don't think we know, so I think a lot still to be done, but how to best get there and answer these questions in a rational way when we have preexisting biases, I think, is where the evidence-based guidelines can help, but I think that sometimes we go a little bit too far in terms of, you know, protecting things we previously did. I think a good example of evidence-based guidelines was for gliomas. I had some initial publications using gliomas. We all thought this was going to be great. We could add as a boost. It was going to be like, you know, brachytherapy seemed to add something that would be better and less invasive. We had a lot of studies that were positive. We even had some class two data that was, you know, positive for radiosurgery, but as people pointed out, these things are really selected. We were selecting patients who responded. We were selecting patients who had smaller tumors. Generally they were younger patients, and when they went to a randomized phase three trial up front, this is a negative trial, and this is not standard of care. It is not practice, and the last time I treated a GBM, it's definitely years ago because there are clearly other better options. So clearly these evidence-type studies are important and should lead to change in practice. This is an example of an evidentiary tables. Again, it just gives an example of one of the GBM studies that it's just the last study that went through, and you do this basically for all the studies and come out with that. I just wanted to make sure that I wasn't, that Antonio wasn't the only person who mentioned Jailburg because this is Flickinger's curve on top of Jailburg's curve. Looking at dose and volume, I think we also know quite a bit about this, and we've learned quite a bit. The RTOG did a very nice study looking at this just in terms of dose-volume relationships, really class one evidence for what dose you can give things based on the overall volumes. Didn't answer the question of what the higher limit was for the smaller lesions, but a good one. I think there's lots of class two to three on others. Benign lesions, I'm going to tell you right now, almost all of it is class three data. There's very little true randomized control data, which is why it gets a little bit harder to answer the fractionation questions Antonio went, and I'm going to get to conclusions because I'm going to get us to lunch closer to time. So the bottom line is that evidence-based methodologies are useful to organize the existing literature and to make yourself ask questions, what do you need to ask next? I think it's very humbling for all of us as we go through and look at this literature and know what we do clinically to realize how little of it is really based on true evidence-based literature. One of my other favorite pet peeves is in glioblastomas, everybody wants to quote what volume you treat, and then, you know, it's the RTOG data. And it's completely, you know, not based on much fact at all. There's, you know, some, you know, some biopsy data. We actually, in our practice, frankly, I don't give margins because I don't really think they add a whole lot to the GBMs, and I've done this now for the last seven to 10 years, and I have a colleague that believes in the RTOG data, and so he follows the RTOG data, and we've looked at our data side by side, and they're superimposable. The GBM patients did exactly the same, and whether you, like, gave essentially a 2-millimeter margin on edema or a 2-centimeter volume, it didn't matter. One principle also with that is that, you know, in general, smaller volumes are better. If you can treat less normal tissue, you should. I think radiosurgery has brought that out and should be high on our mind as radiation oncologists. You wouldn't treat more than you need to, generally. So I'll end it there. Happy to answer questions. Thank you.

evidence-based radiosurgery

methodology

literature review

evidence-based guidelines

brain metastases