What I want to talk about is an integrated decision-making in the management of metastatic spine disease, particularly this framework that we work from at Memorial, and I think sort of regionally and around the country is more being adopted, this neurologic-oncologic mechanical systemic disease framework, and I'll show you how that works. Twenty percent of cancer patients will develop spine metastases over the course of their illness. There's an increased number of spine tumors as MR imaging has improved detection, but really these systemic treatments, including the new biologics, have improved survival, leading to an increased number of spine metastases that we're seeing, and it's very interesting because the newer biologics are clearly getting significantly better systemic control, but they don't work nearly as well in bone as they do for visceral disease. So we see these patients get extraordinarily good control for renal and melanoma, which we never predicted five years ago, and yet we're seeing more spine metastases on the basis of that, which we're now responsible for controlling in the long term. We've had an explosion of new technology, and that's really interesting spine tumors. As Peter mentioned, surgical instrumentation and techniques have improved. Percutaneous cement augmentation for pathologic compression fractures have certainly played a large role, but far and away the biggest advance has been the integration of stereotactic radiosurgery into our treatment paradigms. This is reflected in our multidisciplinary spine tumor center at Memorial, where we went from 40 procedures in 1995 to over 1,000 in 2012, and over 1,200 last year. We've gone from anecdotal institutionally directed treatment to evidence-based guidelines for spinal cord compression, mechanical instability, and now looking at the roles of conventional external beam radiation versus stereotactic radiosurgery typically in our institution has been noted 24 gray single fraction, or eight to 10 gray times three. You have to keep in mind that for metastatic disease, the goal of therapy isn't curative, it is palliative. We want to achieve pain control, improve or maintain neurologic status, achieve local tumor control, and mechanical stability, and ultimately we want to improve quality of life for these patients. Here's a typical case example of a 66-year-old with a history of renal cell cancer, a three-week history of what we call biologic back pain. That's night or morning pain that resolves over the course of the day, and probably has to do with a diurnal variation and endogenous steroid secretion, so when you go to bed at night, you get flare inflammatory pain from mediators secreted by the tumor. As you get up and start walking during the day, your adrenal gland kicks back in, and you start making steroids, and those patients feel better, and these typically respond to exogenous steroids. You can give decadrine to relieve that pain. That's very different from mechanical instability pain, which we'll talk about in a little bit. This patient had the acute onset of weakness, ASIA-C, so less than three out of five in the lower extremities, chronic renal insufficiency, and the systemic workup showed renal cell extending the renal vein, some pulmonary nodules, and an acetabular fracture, and you can see here this extraordinarily high-grade spinal cord compressor with circumferential disease, renal cell carcinoma, and I think 10 years ago, we had really limited options for this patient. Chemoimmune or hormonal therapy certainly didn't have great effect in renal cell, so we really had conventional external beam radiation versus surgery, and even at that point, we had anterior, posterior, lateral, or combined. In 2014, we now have, again, the introduction of these targeted therapies, even for renal cells such as SUTENT and EXAVAR, but again, very limited responses in bone. We have conventional external beam radiation, but again, the introduction of image-guided intensity modulated radiation therapy, giving you the hypo-fractionated regimen eight to 10 grade times for your high-dose single fraction, have fundamentally changed the responses in these tumors and need to be integrated into our approach, and now we even have brachytherapy plaques such as P32, these high-dose rate beta-emitting radioisotopes for dural recurrences or to treat dural margins, and then high-dose rate brachytherapy, iridium after loaded catheters to treat bone disease or bone recurrences, and again, from surgery, we've added percutaneous cement augmentation to our armamentarium of how we would treat a patient like this. So for a high-grade circumferential renal cell carcinoma in this setting, how many residents would treat this patient with upfront radiation, either conventional radiation or stereotactic radiosurgery? So Sam didn't convince much on that one, but, and how many would go for open surgery initially? Good, and if you went for open surgery, would you follow up with conventional external beam radiation? How many people would give 30 grade and 10? Okay, and how many people would go for radiosurgery as opposed to operative adjuvant? Good. And I'll try to show you the data on how we got there. When I got to Memorial almost 20 years ago, I literally had no idea how to make decisions about these patients, and so we really came up with not an algorithm, but a decision framework that we referred to as NOMES, neurologic, oncologic, mechanical, stability, and systemic disease to decide best treatment, systemic, radiation, or surgery, and really, it was just a way to organize our thoughts into what we thought were the four most fundamental assessments that we needed to make in this patient population. From a neurologic perspective, we concerned about myelopathy and functional radiculopathy, but for decision-making purposes, we're principally concerned about the degree of spinal cord compression. From an oncologic perspective, we worry about the tumor histology as it relates to the radiation and our chemosensitivity of the tumor. We're separately worried about, is that patient stable or unstable? Why? Because no amount of radiation therapy will stabilize an unstable spine, so once you determine instability, there needs to be some kind of procedure to stabilize. And finally, what are the systemic disease and medical comorbidities that impact on our ability to either offer surgery or radiation therapy? The neurologic and oncologic decisions are made in combination. From an oncologic perspective, again, what we're principally concerned about is the radiation sensitivity of the tumor, here to conventional external beam radiation, so there's sensitive and moderately sensitive tumors, such as myeloma, lymphoma, breast and prostate cancer, and then there are moderately to highly resistant tumors to 30 gray and 10 fractions, such as colon non-small cell lung cancer, thyroid, renal, most of the sarcomas, and melanoma. And then from a neurologic perspective, we look at the degree of epidural spinal cord compression. So we're very concerned about whether a patient has myelopathy or functional radiculopathy, but for decision-making purposes, what we're principally concerned about is the degree of cord compression, and there's a scoring system, as Peter showed earlier, where zero is bone only, one epidural impingement, two spinal cord compression but CSF is seen, three spinal cord compression, no CSF seen, and for decision-making purposes, the twos and threes are considered to be high-grade spinal cord compression. So in 2014, how do we make decisions? Well, for the sensitive and moderately sensitive tumors, regardless of the degree of spinal cord compression, we're very comfortable at rating those patients 30 gray and 10 fractions with the expectation that the tumor will apoptose and decompress the spinal cord. For patients with moderately to highly resistant tumors with bone only or epidural impingement, we've seen very poor responses to 30 gray and 10 fractions, and so for these patients, we typically take them for upfront stereotactic radiosurgery, again, in our institution, typically 24 gray single fraction, and for moderately to highly resistant tumors with high-grade spinal cord compression, it's very hard to conform the beams tightly enough to spare spinal cord tolerance using stereotactic radiosurgery, and so for these patients, we upfront treat them with a simple decompressive surgery called separation surgery and then followed up with stereotactic radiosurgery to get local tumor control. How do we know about the radiosensitivity based on conventional external beam radiation and why do we treat, with conventional external, those patients who are radiosensitive regardless of the degree of spinal cord compression? Well, there's very little in the literature that actually looked at radiosensitivity based on differential tumor histology, but the ones that did were remarkably consistent. Lymphoma, myeloma semino, breast and prostate cancer respond dramatically well to 30 gray and 10 fractions versus those, really the remainder of the solid tumors, are remarkably radio-resistant. The radiosensitive tumors have about an 80% response rate at 16 months. The radio-resistant tumors have about a 20% response at 3 months to conventional radiation, which really begs the question of whether they're getting a response at all. Here's a patient with multiple myeloma T10 with very high-grade spinal cord compression. Patient got 300 centigrade times 8 and insisted on a reimage, and you can see the tumor's completely decompressed after a very small amount of radiation. So for the radiosensitive tumor, it's giving conventional external beam radiation is a very effective modality for treating the tumor and decompressing the spinal cord. Unfortunately, conventional external beam radiation, we very rarely see these responses with the radio-resistant tumors, and for that reason, we've really gone to stereotactic radiosurgery to treat tumors up front setting with minimal or no spinal cord compression. There are obviously a number of platforms that can deliver this kind of radiation. It's very high-dose conformal photon therapy. Here's a series that Josh was talking about earlier, and everybody who's presented here on spine radiosurgery has a similar series. This was 413 patients. It was a dose escalation from 18 to 24 gray. Our three-year recurrence rates overall were 4%. 18 to 23, 10% was a dose response. Those patients who got 24 gray had a 2.4% response rate at three years, and for us watching these patients go through this, this is an absolute game-changer. We're going now from about 20% response at three months, or 80% recurrence at three months, to 2.4% recurrence at three years, and you can imagine what that does for a patient population that you're trying to palliate who's living significantly longer on these newer biologics. What was most compelling was it was histology independent. It didn't matter whether it was breast or prostate cancer versus renal melanoma, sarcomas or thyroid. They all had greater than 90% response, and again, that 24 gray line had over 95% long-term durable tumor control, and again, in this patient population, that is extraordinary. Josh talked about the toxicity. We see a lot of grade 1 and 2 skin and esophageal. We talked about the esophageal perforations. We've had three cases of myelopathy, mostly radiographic, one of which was suboptimally treated. There's a question of whether we're seeing radiculopathy and plexopathy, which we think we do have an incidence of this, mostly in watershed zones, particularly around the sciatic nerve, and some of these may be related to having VEGF inhibitors, which may potentiate the effects of radiation in these patients. We talked about the vertebral body fractures, and I won't go through that other than to reiterate that we do have a 40% progressive, our new fracture rate radiographically, 10% of which are symptomatic, and it is fairly straightforward for us to treat those symptomatic fractures when they occur with vertebral kyphoplasty. So a strong recommendation was made by the Spine Oncology Study Group. This was published in the Spine 3422 supplement, with low-quality evidence that radiosurgery should be considered over conventional fractionated radiotherapy for the treatment of solid tumor spine metastases in the setting of oligometastatic and or radio-resistant tumor histologies in which no relative contraindication exists. So how has it changed our treatment paradigms? And again, I think Peter showed this earlier, but here's a patient with renal cell carcinoma with T10 disease, and you can see here there's this little bit of epidural impingement, paraspinal mass in this vertebral body tumor. If you went to old frameworks for how to treat this, this is the Tamida and Takahashi scores were very popular. If you score this based on the Tamida scoring system, this is a moderate growth tumor, solitary isolator for long-term local tumor control. The recommendation was wider marginal excision, so you're going to en bloc excise this in order to get local tumor control. This is based in the published literature on six case series of a total of 15 patients. Operative times were 8 to 12 hours. The transfusion data was reported in a single series, 16 units packed red blood cells and 20 units of FFP, and despite that transfusion data, there were no complications reported. Recurrence rates were 13 percent. Most of these patients, however, followed with plain x-rays, and the median follow-up was only 16 months. If you look at our radiosurgery series of 80 patients looking at renal cell and melanoma over a four-year period, dose escalation from 18 to 24 gray, imaging and physical exam every four months, radiographic and symptom control were 92 percent with a trend towards better control at 24 gray, 97 percent local control with renal cell carcinoma in the same setting that they were treating these patients aggressively with en bloc resection. So if we look at our NOMS assessment for renal cell carcinoma with an ESCC score of one, so epidural impingement but no spinal cord compression, instead of being resistant to conventional external beam radiation, now we're exquisitely sensitive to radiosurgery. The patient's stable, which we'll talk about in a minute, and this patient's best served with stereotactic radiosurgery. You can see here that the patient, the epidural disease disappeared, the paraspinal mass is gone, and you're always left with these bone lesions, which we'll talk about in a minute, but this patient has extraordinarily good control at 26 months out. Now, there's been an argument that if you took out solitary metastases from renal cell that you could cure the patient. In fact, the median time to progression, even with en bloc excision, is 11 months, and at 26 months we already have metastatic disease. Again, there's no intent to cure with stage four metastatic disease, and our best intent is to palliate. So a strong recommendation is made that patients with solid renal carcinoma in the absence of epidural disease may benefit from stereotactic radiosurgery as first-line therapy rather than en bloc excision. The question, again, has been, what do you do with these bone lesions? There's a lot of work trying to determine whether this is actually a treated lesion or not. It turns out that PET scans usually turn sort of lukewarm, so it's hard to figure out whether this is actually a treated disease, but it turns out that we use dynamic contrast imaging now, DCE, particularly looking at the plasma volumes are very predictive of active tumor recurrence long before standard MR imaging, and so we're trying to employ this on many of our patients to see whether we're actually getting a significant response from that bone lesion on MR that will never really change or go away, and this has been remarkably effective. Unfortunately, we can't use it in the post-operative setting because you can't do the computer algorithm based on that, but certainly in the up-front setting, we have a very good sense of whether these are treated or not. So we've talked about radiosensitive tumors regardless of the degree of cord compression we're going to treat with conventional external beam radiation. For radio-resistant tumors with minimal or no spinal cord compression radiosurgery, how do we justify taking patients with radio-resistant tumors, high-grade cord compression for surgery followed by stereotactic radiosurgery? Well, it turns out, at least in Memorial, the seven local failures we saw all received less than 15 gray to some portion of the planning target volume. If we have a cord maximum dose, point dose to less than 14 gray with a 10 percent per millimeter fall-off, then either we're going to give a subtherapeutic dose to the area around the spinal cord and risk epidural progression where we need the greatest control, or we're going to give a cytotoxic tumoral dose at the margin of the cord and ultimately hurt the spinal cord. And so, at least in our institution, we don't treat high-grade spinal cord compression with radio-resistant tumor, with radiosurgery. And I really admire and really have a great deal of faith that ultimately we may get there. And Sam has done a remarkable job at trying to push that agenda. But I think at the current time, it's really not a standard practice. And if it's on protocol, it's probably reasonable. But as we stand now, it's probably not what we should be doing, certainly in the community and certainly off protocol. The caveat, I think, is what he showed. I think in a breast cancer patient, you probably can get away with a lower dose and spare spinal cord tolerance and still have very good effect. But for the radio-resistant tumors, even in his series, it wasn't so much a matter of not getting people back to ambulation. But half of the patients fell out in that series. And then nine patients actually progressed neurologically. And I think that's where we run into problems, is in an institution or in a community where you have surgery available, are you going to allow people to progress on radiosurgery? And it may be too high of a price to pay currently until we have better data on that. But I admire the work that's been done. And I think for now, we really are sort of obligated to follow the PATCHEL study, which is a remarkably good study, randomized controlled trial, looking at solid tumors, high-grade cord compression with myelopathy, comparing surgery plus conventional external beam radiation to conventional external beam radiation alone. And again, rightly so, they excluded the radiosensitive tumors. Why? Because if they included it, it would have been very heavily weighted towards the radiation group. And in terms of, and Peter showed this data, but overall rates of ambulation were better in the surgical versus the radiation group. The duration of ambulation was until death in the surgical group versus only 13 days in the radiation group. The duration of ambulation was 62% in the surgical group versus 19% in the radiation group, but it was an intention to treat analysis. So all three patients who ambulated in the radiation group were actually crossover patients to the surgical group. Condoms was maintained longer in the surgical group. Narcotic requirements were actually lower in the surgical versus the radiation group. And survival times were statistically significantly better in the surgical versus the radiation group. And based on this, a strong recommendation is made for patients with high-grade cord compression due to solid tumor malignancy undergo surgical decompression stabilization followed by radiation therapy. The question became what kind of radiation therapy and how much surgery did you really need to get patients there? And here's an 86-year-old. She's got papillary thyroid cancer, so a massively hypervascular tumor, even embolized, it's going to bleed. She's age is C, so she's less than three out of five in her low extremities and already lost proprioception. And you can see here she's got very high-grade spinal cord compression, three-level vertebral body disease. And there's no way on this planet that you're going to get an 86-year-old through a three-level front-back surgery, that you're going to be able to take out all that disease, get her through and do something meaningful for her. But when we used conventional external beam radiation as opposed to operative adjuvant, we had no faith that it would actually control that tumor. And so we were very aggressive with surgical resection, and we probably introduced a lot of morbidity based on trying to get tumor control with surgery. With the evolution of stereotactic radiosurgery, we're much more comfortable simply doing a decompression of the fecal sac. All we really have to do is reconstitute the fecal sac, do a long posterior segmental fixation and compression, and then treat the remainder of that tumor with radiosurgery. Why? Because the responses are dramatic, consistent, and volume-independent. And so for this patient, we actually did a very small decompressive surgery. We left all that vertebral body tumor alone. We reconstituted the fecal sac. Here's the postoperative myelogram for treatment planning purposes. We did a long posterior submental fixation and compression, got this patient, in this case, to hypofractionator radiation because it was three segments, nine grade times three, two years out, fully ambulatory, well-controlled disease. We gave this patient a surgery that she could tolerate and got her to effective radiation for tumor control. What do we know about conventional external beam radiation as opposed to operative adjuvant? This is basically the only series that ever followed patients longitudinally out. It was Clay-Kamp-Sammey published in 1998, 101 patients, mostly operated through a post-stelotype resection with a relatively aggressive resection of the tumor. The adjuvant there was predominantly conventional external beam radiation, and the local control was only 30 percent of the year, and if the patients lived long enough, all of them recurred. What was the most significant predictor of recurrence tumor histology? Why? Because conventional external beam radiation for those tumors works no better in the post-operative adjuvant setting as it does in the up-front setting. Here's our series of using radiosurgery as a post-operative adjuvant, and again, we're not doing these massively aggressive resections. We're simply doing what we call separation surgery. We decompress the fecal sac, long posterior fixation, 186 patients. There were three strategies, post-operative adjuvant, HIDO single fraction, 24 gray, HIDO hypofractionated, eight to 10 gray times three, and then low-dose hypofractionated, six gray times five. And our one-year cumulative incidence of recurrence was 16.4 percent, but in the HIDO group, single fraction or HIDO hypofractionated, we were less than 10 percent recurrence rate. Dose did matter in this with the low-dose hypofractionated doing significantly worse. There were no neurologic complications from this and no association with radio-resistant tumor histologies. Again, these responses are histology independent. Prior radiation, 50 percent of these patients had already failed conventional external beam radiation before they progressed with high-grade epidural disease requiring surgery or the degree of epidural extension preoperatively. The criticism we've gotten for this is that if we don't take out the vertebral body tumor or reconstruct the anterior column, these constructs are going to fall apart. And so we went back to look at that in 319 patients and our local failure rate, our failure rate of our construct was only 2.8 percent. Why? It's not because these patients, which is significantly better actually than some of the other series like degenerative disease, the issue with these patients is they're not going back to high-impact activities. We're getting them back to ADLs and they really don't survive long enough necessarily to break these constructs. The exception has actually been renal cell with intact vertebral bodies, thoracolumbar junction. For some reason, we see a ton of them. And truthfully, their survival was a year to two years with that disease 10 years ago. Now they're surviving. We have multiple patients out five years because of the new biologics. And now we're beginning to see broken rods in that patient population. And I actually replaced six rods this year. Why? Because the radiosurgery is so effective, but we're still getting the same fractures in the midst of our construct as we are in the upfront setting. And so they'll fracture that bone and then they'll start putting a lot of stress on their hardware. And we tried to figure out a number of strategies to fix that because it's a new problem for us. We looked at cobalt chromium rods, which are stiffer and stronger. The problem is there's a lot of shielding from the radiation, so you can't put them up in the upfront setting. But now we image these patients, and as soon as you start to see end plate changes, we'll do kypho vertebroplasty on those patients around the construct, and that may be a better strategy for us ultimately. So we go back to this initial patient who had renal cell carcinoma. This is using our NOMS framework. The patient's myelopathic, ESCC score of three. Renal cell is markedly resistant to conventional external beam radiation, which at least in our institution is the only radiation you can use in a setting of high-grade cord compression. This patient's stable, and based on our NOMS framework, we would put this patient on hydrosteroids, embolize the patient, and then take him for this very simple separation surgery, which takes about two hours, and then get him to post-operative stereotactic radiosurgery to get local tumor control. I just want to touch both for the surgeons and really for the radiation oncologists the concept that Peter touched on of mechanical instability. Remember, there's no amount of radiation that's going to stabilize an unstable spine, and you may see as a radiation oncologist that patient before the surgeon does, and in some way you have to make a determination about whether that patient is stable or unstable. And the Spine Oncology Study Group came up with a scoring system called SINS, the Spine Instability Neoplastic Score, which at least gets you into the ballpark of who's stable or unstable. It is reliable and valid, but it's a little bit more difficult, I think, because it's so radiographic heavy in its assessment for radiation oncologists to actually use. But there are six components to this, and each is weighted by their contribution to instability. So location, junctional, and mobile spine have a significantly higher point score than semi-rigid or rigid spine. The pain is probably the most important thing, especially for the radiation oncologist. Again, there's this real differential between biologic pain, night or morning pain, which resolves over the course of the day. And truly 95 percent of the patients I see have that kind of pain. You say, when do you hurt? They say, I go to bed at night, and I wake up in the middle of the night and I'm in agony. And when I get up and start moving during the day, I feel better. That's not instability pain. That doesn't get any contribution to instability. That's just the tumor being there with these inflammatory mediators. Anytime you hear a story of, I get up and move, I have axial load pain, I bear weight, I sit or stand and I get back pain. That deserves some kind of assessment by a surgeon or somebody who knows about instability to make that determination. I think counterintuitively for us in the thoracic spine, it's very interesting, but it's not really flexion pain or axial load pain, but it's lying flat pain. Because often they're kyphotic, so they're in this position. When they lie flat, they're hinging on an unstable kyphosis. That's the patient who comes in and says, I won't get flat in bed. I've been sitting in a recliner for the last three weeks because I can't get flat in bed. When you hear that story, think about instability in the thoracic spine. For the radiation oncologist, I think what you hear in the clinic is probably more important than a radiographic assessment, which again, I'll show you, is a little bit duff to make. Bone lesions were assessed as lytic, mixed, or blastic, alignment, subluxation, translation got the highest point score, the degree of vertebral body involvement, and then post-treatment involvement, and those scores are tallied, zero to six was stable, seven to 12 potentially unstable, and 13 to 18 unstable. So here's a 45-year-old woman with hormone refractory breast carcinoma, right? So you do your NO assessment, no matter what compression they have, breast cancer's going to be markedly radiosensitive, but the patient has progressive neck pain on flexion, extension, and lateral rotation, and has severe occipital neuralgia. Imaging shows this, and you can see here, if you look at her posterior laminar lines, here's C2, and C1 has slipped all the way forward. So the patient has a significant fracture subluxation, significant lytic bone destruction based on MR scan. So if we score this patient, it's a disjunctional spine, significant movement-related pain, lytic bone destruction, significant subluxation, translation, less than 50% for vertebral body collapse, but both postural elements are involved, SYN score is 17, grossly unstable. Again, NO, remarkably sensitive to radiation, why did we take this patient to surgery? Grossly unstable. And this patient underwent a C1-2 laminectomy with occipito-cervical fixation. Here's another case example, and this is the newest that we're doing for instability. 38-year-old, had non-small cell lung cancer, metastatic of spine, got conventional external beam radiation, which typically today we probably would have treated this patient with SRS, T12L1, and then had an acute onset of fracture at T12L1. So this bone fractured, and the patient was neurologically intact. We talked about the complications of operating on patients, but if you give conventional external beam radiation and have to operate within six weeks, there's a 50% risk of wound dehiscence. And so despite this instability, thoracolumbar junction, we didn't want to take this patient to surgery. We scored this patient grossly unstable, and this patient's a very good candidate for percutaneous pedicle screws. So we like chyfome vertebroplasty, but in the setting of recent irradiation, in the setting of instability, or the patient has a burst fracture with posterior element disease, they need a posterior tension band. And these can be very well done just through little percutaneous incisions, the screws are put in. We cement augment the screws to make sure that we're stable, and these patients, I think we now have about 60 of them, have done remarkably well. The last assessment you have to make is really the systemic, and both an extended disease and a medical workup to make sure that whatever you propose, that patient can tolerate in the context of palliation. So here's a patient with undifferentiated sarcoma, 18 centimeter paraspinal mass, age is C, less than three out of five in the lower extremities. This is a really simple separation surgery SRS case. This is an hour and a half, we can get this patient decompressed. The problem is that they had IVC clot extending to the right atrium, and they said just by turning that patient prone, we had a 50% chance of mortality, and that we weren't willing to take. This is palliative. This patient can have great quality of life, but if we take them and we harm them in surgery, it's not worth anything to them, and so this patient became paralyzed, but lived for whatever time they had with their family, doing well. The other thing that we learned, I think, is what epidural disease means in the context of some of the tumor histologies. So this was just a 57-year-old small cell bladder cancer, was multiply operated, got a spine metastasis, was irradiated 30 grade, and 10 recurred, got irradiated six grade times five, and then had progression of epidural disease, but the extended disease workup was negative, so they said take this disease out and bring it to negative disease, and we took that little bit of disease out, and within six days of a negative extended disease workup, the LFTs went up, and the whole left lobe of the liver was replaced with tumor. Why? Because sometimes epidural disease is a harbinger of explosive systemic disease that's about to happen, and it's actually remarkably predictable. The bad actors, probably the worst, is hormone refractory prostate cancer, non-small cell lung cancer, colon cancer, and hepatocellular. When they get epidural disease, they're in a bad place systemically, and the first place you see it may be epidural, but they're going to explode oftentimes, and so you need to at least consider that when you talk about taking a patient to surgery, or even stereotactic radiosurgery. This was a very old series of ours, 140 patients that we did postulal resections on, and we had the sense that those were our bad actors, but in fact they were. The survivals for colon, non-small cell lung, and hormone refractory prostate were less than six months. Why? Not because you couldn't get them through surgery, but because they were about to explode, and you just caught it in the epidural space first. So overall, for patients who are radio-sensitive tumors to conventional external beam radiation regardless of degree of cord compression, we're very comfortable taking them to conventional external beam radiation on high-dose steroids. For radio-resistant tumors with minimal or no spinal cord compression in our institution, we go straight to stereotactic radiosurgery. For radio-resistant tumors with high-grade spinal cord compression, we do this very simple separation surgery. We reconstitute the fecal sac, long posterior fixation, but no intent to take out all the tumor, and then follow them up with stereotactic radiosurgery. If they're unstable, they need some kind of stabilization procedure, either kypho-vertebroplasty or percutaneous pedicle screws, and everything is predicated on what they can tolerate from a systemic disease standpoint. Thank you.

metastatic spine disease

neurologic-oncologic mechanical systemic

stereotactic radiosurgery

pain control

local tumor control

tumor histology

individualized treatment plans