We'll speak on on-block resection of primary spinal tumors. I'd echo what the previous speakers have said and say thank you, Wash, for having me. Okay, we're going to switch gears fairly considerably now and talk about primary bone tumors and on-block resection. First point I want to make is that compared to what you've just seen, primary bone tumors of the spine are rare. Even in a major tertiary referral center like MD Anderson, primary bone tumors only represent approximately 10% of our work, and in the community they're even less common. So my initial message to you as surgeons is have an index of suspicion and recognize the possibility that the patient you're treating may have a primary bone tumor, because as you'll see from this talk, they're managed quite differently from metastatic disease, and if you treat a primary bone tumor like a metastasis, you can perform an operation that will take a potentially curable patient and render them incurable. It is not only important to recognize a primary bone tumor when you see one, but it's also important to recognize what type, just like in the case of metastatic disease. These range from the benign to the malignant. Why is it important to know what type? Well, because the biological behavior of the tumor and its grade will influence things like survival and how aggressive a surgery you need to manage these tumors. Why else is it important to know what you're treating? Because different primary bone tumors will respond differently to chemotherapy and radiation. Tumors like Ewing sarcoma, osteogenic sarcoma, now giant cell tumor, the first stop for those patients is typically with the medical oncologist, whereas tumors like chordoma and chondrosarcoma tend to be relatively resistant to radiation and chemotherapy and are primarily surgical diseases. Finally, it's important to know what you're treating, because some of these tumors are extremely hypervascular and might benefit from a preoperative embolization. Hemangioma, aneurysmal bone cyst, giant cell tumor, any tumor with angio or hemangio in the prefix or suffix is likely to benefit from embolization. So you want to know what you're treating, and this message is the same as with the metastatic disease. You remember of the NAMS criteria, O, oncologic, the primary tumor type, is vital to understanding how to manage these patients. Once you've recognized a primary bone tumor and identified it, and we're talking about surgery, it's important to recognize that any surgery that you do is predicated on the goals of your surgery. Are you trying to palliate the patient? Are you trying to cure them? This will influence the type of surgery and magnitude of the surgery that you do. So as you've heard in the last couple of talks, for metastatic tumors, our goal is palliation. Likelihood of distant metastasis is high. Adjuvant therapies are commonly available, and the resection is generally intralesional, delicately painting the tumor away from the fecal sac, performing a separation surgery or a LIT, for example. For primary tumors, however, our goal is different. Here we're trying to cure or create prolonged disease-free interval. The likelihood of distant metastasis is often low. Adjuvant therapies are often limited, like for chordoma and chondrosarcoma. And here we're thinking about an en bloc resection, if feasible, taking the tumor out all in one piece. This is what an en bloc resection looks like. It's an attempt to remove the tumor all in one piece with a layer of normal completely around the tumor. Now, you're looking at this picture and the slide I showed two ago, and anyone would recognize that this is a much bigger operation and much more morbid, and so the onus is on me to prove to you that this is necessary for the treatment of these tumors, and so I'm going to attempt to do that next. Why is en bloc resection necessary? Well, it's based on two lines of evidence, both indirect and direct. The indirect evidence is what we've learned from our orthopedic colleagues treating these tumors in the long bones, where these tumors had a high propensity for local recurrence, typically due to incomplete surgical resection by intralesional techniques. If en bloc resection is better in the appendicular skeleton, it stands to reason the same may be true in the spine. This, of course, is the work of Enneking, who looked at these primary bone tumors in the appendicular skeleton and, based on their biological behavior, determined the nature of surgery required. For aggressive, benign, and primary malignant bone tumors without metastasis, en bloc resection was felt to be better outside the spine, so it stands to reason that the same may be true in the spine. So what happens when we apply Enneking's principles to the spine? The evolution of the data began with small, single-institution case series. Here's Boreani's series of 50 years' experience of chordoma of the mobile spine, and you'll see that when an intralesional excision was performed, there was almost uniform local recurrence, but when he got an en bloc resection, the recurrence rates were markedly reduced. Same is true for our experience at MD Anderson with sacral chordoma. How you do the surgery matters. When you do an en bloc resection, your disease-free survival is better than when you don't. Same true for Boreani's series of chondrosarcoma. Higher rates of local recurrence with piecemeal excision, but if you get an en bloc resection with clean margins, you can markedly reduce the rate of local recurrence, and we've seen the same results in our series at MD Anderson. This has been further expanded in recent years by performing large, systematic reviews, a combination of a review of the world's literature with expert opinion, and our reviews by the Spine Oncology Study Group of chordoma and chondrosarcoma have shown, with strong recommendation based on moderate quality evidence, that for these diseases, en bloc resection is appropriate as it both decreases local recurrence and improves survival. Here is the data. The chordomas are in the upper two panels, chondrosarcoma in the lower two panels, and you'll see that the data from the systematic review showed markedly increased rates of local recurrence when an intralesional resection was performed, and this was also associated with worse survival. Same is true for chondrosarcoma of the spine. More recently, the AO Tumor Knowledge Forum looked at large, international, multi-institutional retrospective series, the largest series ever reported for chordoma and chondrosarcoma of the spine, and you can see again that en bloc resection and anything appropriate resection will significantly reduce the rates of local recurrence for both of these diseases. The graphs are shown here, 166 patients with mobile spine chordoma, 111 patients with mobile spine chondrosarcoma. These results even hold true for the aggressive benign tumors, the stage 3 osteoblastomas and giant cell tumors. Our systematic review showed that when feasible based on staging, en bloc resection seemed to have lower rates of local recurrence, and this has actually recently been shown in retrospective studies from the AO Tumor Knowledge Forum to hold up. We looked at 82 patients operated on for spinal giant cell tumor. The pathologic specimens of the resected specimens were reviewed, and the intralesional margin, a contaminated margin, was much more highly associated with local recurrence, and if local recurrence happened, it was highly associated with death. So I think I've shown you substantial evidence to support the role of en bloc resection in the management of aggressive, benign, and primarily bone tumors in the spine, but saying it's a good idea and showing you all this data and actually carrying out these surgeries safely may be two different things, and so I want to emphasize here that if you're going to take on these cases, you need to have several things. You need to be able to plan carefully. You need to have multidisciplinary surgical and oncologic expertise in your hospital, and you need to have the latest technology. When we see these cases, there are important things that we consider before taking the patient to the operating room to establish that the operation is technically feasible, and these are very important, and we review them in every case. Can we come up with a strategy for en bloc resection? Can we open the bony ring of the spine in areas of normal such that we can free the neurological elements and release the diseased portion of the bone all in one piece? Number two, can we achieve an adequate exposure and approach to allow us to achieve this objective, and do we need assistance from surgeons in other disciplines to help us make this happen safely? Can we rebuild the spine once we're done, and have we thought of a strategy for soft tissue coverage and adequate healing? At the end of the day, these patients need a durable benefit with satisfactory postoperative function and acceptable morbidity. For every case, we plan, plan, plan. We look for multidisciplinary surgical expertise that can help us, and we consider whether there's technology that can also work to our advantage. We actually use this WBB surgical staging system, which divides the spine into 12 radiating zones in five concentric layers. We superimpose the tumor on this picture, and we figure out preoperatively where we're going to cut the bone in area of normal to free the neurological elements and release the diseased vertebra all in one piece. So I thought I'd spend the next few minutes showing a couple of example cases to show that this is actually feasible with appropriate planning. First is a 45-year-old female who presents with neck pain. MRI shows a mass at C4. Chest, abdomen, and pelvis are negative, and CT-guided biopsy revealed chordoma, patient's neurologically intact. You can see here this lesion involving the C4 vertebra with this prespinal retropharyngeal component shown on the MRI. Here is the same case shown in CT. You can see the erosion of the bone, and there's the faint image of the prespinal mass. We immediately look to our WBB diagram to figure out where are we going to open this bony ring where there's no tumor so that we can free the neurological elements and remove the diseased portion of the tumor all in one piece. For tumors in the vertebral body, this frequently involves going through the pedicles. Not easy in the cervical spine due to the vertebral arteries, but you can perform a medial facetectomy, enter the foramen transversarium, push the vertebral artery laterally, come through the front of the foramen transversarium, and effectively release this diseased vertebra. We perform this via a two-stage approach, the posterior operation for the release of the neurological elements and release of the vertebra by cutting the bone, followed by a spinal stabilization. And we follow this with a second-stage anterior approach where we cut the disc above, cut the disc below, and remove this vertebra all in one piece. Would help from another surgical discipline assist us with this? You're right. Head and neck surgery can be used to help us elevate the posterior pharynx, elevate the esophagus and airway away from this large prevertebral tumor mass. So let's actually show you the intraoperative pictures. Head is up here. Legs are down here. You can see the C4 laminectomy has been performed, and a medial facetectomy has been performed as well, revealing the exiting C4 nerve roots. Here's the blown-up view from right here, and you can see immediately below the C4 nerve root is the C4 pedicle. The pedicle has been removed here. The foramen transversarium is entered. The vertebral artery is pushed laterally, and this osteotome is then passed through the front of the foramen transversarium, releasing the nerves, freeing the diseased vertebra from the normal adjacent bone. We then stabilize the spine, and here's a CT scan done after the stage 1 operation. You can see now the spine has been stabilized, and this diseased vertebra has now been isolated. Now theoretically, all we have to do is come in from the front, cut the disc above, cut the disc below, and this vertebra should be free, and this is exactly what we did. Here's the left-sided neck approach done by our head and neck surgeons. Head is up here. Legs are down here. Pharynx, esophagus are elevated away from the tumor, and we're able to expose the spine above and below, C3 and C5. Careful exposure of the disc and complete removal of the disc and annulus at C4-5 and C2-3 allow this vertebra to come completely free. You can see the fecal sac in the background, and we can remove this vertebra along with the associated tumor all in one piece. This is the pathology specimen in tight correlation with the preoperative MRI. The spine is reconstructed, and this patient went on to have proton beam radiotherapy after the surgery and remains disease-free eight years after the surgery, fully functional. She's an older woman who takes care of her grandkids while her daughter works and has a very good functional result. After surgery, some postoperative swallowing difficulties which resolved within three weeks of the surgery. Just to show you another case, how we adapt our planning to what the tumor gives us, 35-year-old man with pain and numbness in the right chest. CT scan shows an intrathoracic mass. Biopsy reveals a giant cell tumor. He was treated at that time about eight or nine years ago with interferon and serial embolizations, and then was referred to us for definitive surgical resection. Here is this tumor involving the T4, T5, T6, and 7 levels. The tumor is primarily involving the chest wall but adjacent to the spine at T5, T6, and T7. So again, we don't violate our principles. We use our WBB diagram to figure out where we're going to cut in areas of normal such that we can get this tumor out all in one piece. In this case, this involves a cut through the chest wall, lateral to the tumor, and a sagittal osteotomy through the vertebra, medial to the tumor. We also do this surgery in two stages. First, a posterior approach to open up the lamina, expose the fecal sac, ligate the nerves on the right and initiate our osteotomy, followed by a stabilization. And then we come back in a second stage where we put the patient with their right side up, reopen the back, and perform a simultaneous thoracotomy so we can see to the front and the back simultaneously, passing our osteotomes from back to front as safely as possible. Will we need other surgeons to help us with this case? Absolutely. I'm sure you've figured out that a thoracic surgeon is going to be valuable to A, help get us in the chest, B, help mobilize the vessels to protect the vessels from when the osteotome comes through, and C, to free any diseased lung and leave a nice margin along the tumor. Any other surgeons that might help us? How about the defect that's going to be left behind when this tumor comes out? We can get our plastic surgeons involved at the beginning of the case, and they can harvest soft tissues to help us free the defect. Anything else that might help us with the planning? This is not an easy osteotomy. Keep in mind that even with the back and front exposed simultaneously, there's a big tumor overhang. It's going to be hard to see that your osteotomes are going in exactly the right direction, and there's not a lot of room for error. Too far to the left, and you're in the aorta. Too far to the right, and you're in the tumor. How can I ensure that I'm making those bone cuts in the right place? Maybe we can use our intraoperative navigation to help us make these osteotomies safely. So let me show you how that's all done. Here is the stage one procedure. The patient is prone. Head is up here. Legs are down here. The laminectomy has been performed. The tumor is sitting over here in the chest wall. Via the lamina, we're able to get into the right epidural space and sequentially sacrifice the right-sided nerve roots. These are going to the chest wall that's going to be removed. With the nerve roots sacrificed, we can get access to the back of the vertebra, immediately medial to the pedicle, and with a one-millimeter cutting burr, we can initiate our osteotomy at the T5, T6, and T7 vertebra, giving us a docking area for our osteotomes and a mark on the spine that will allow us to confirm the accuracy of our subsequent navigation. After this, the spine is stabilized, and here's a CT scan done after the stage one procedure but before the stage two procedure. Spine has been stabilized. Tumor is sitting here, and you can see the initiation of our osteotomy docking zone. Now we bring the patient back for stage two. Head is up here. Legs are down here. We're going to reopen the spine incision and combine this with a posterior lateral thoracotomy. On the way in, our plastic surgeons move the latissimus, trapezius, and serratus muscles out of the way, out of the damaged path of the surgery, so these can be preserved for the subsequent soft tissue reconstruction. We're thinking about this in advance. Then with the help of our thoracic surgeons, we reopen the back, cut the fifth, sixth, and seventh ribs distally, and gain simultaneous access in the chest. You can see in the back the paraspinal muscle has been truncated. Let me roll you into the chest and show you why it's so good to have the thoracic surgeons. Number one, hypervascular tumor. The lung's been parasitized. We need to take a wedge resection of the lung to protect the margin here, and we can also use our thoracic surgeons to mobilize the aorta away from the front of the spine and actually put their hand in front of the spine to protect those vessels during the subsequent osteotomy. Much better to have a thoracic surgeon's fingers in the path of the osteotome than a valuable neurosurgeon's fingers, so we use our thoracic surgeons for that. And this is very, very, very beneficial. So we have the thoracic surgeons having mobilized everything in the front, and now we have to make our bone cuts. We hold the muscle out of the way. We remove the rod. We relocate our osteotomy docking zone, and now we apply our image guidance. We link, put our reference array on the spine at the base of our construct. Here's our pointer probe checking the accuracy. The probe is in the predrilled osteotomy defect, and you can see on the image guidance that we're precisely accurate. The probe is in the osteotome docking point. Now we can actually put a reference array on the actual osteotome, and with real-time image guidance, tap through from back to front, putting our osteotomy exactly where we planned it to be, safely away from the aorta and safely away from the tumor. This allows us to cut the vertebra in the sagittal plane. You can see the fecal sac in the foreground with the ligated nerve roots. Here's a better view of the cut vertebra, the part going with the tumor, the part that gets left behind. This allows us to remove the tumor all in one piece. Here's the en bloc resected specimen intact with no contamination. Here is the reconstruction of the spine, courtesy of our plastic surgeons. We get a very nice cosmetic result. We reduce the risk of wound healing complications, all because we thought and planned ahead. And this is the postoperative CT scan with the osteotomy exactly where we planned it to be and soft tissue filling the defect. Satisfactory reconstruction. This patient did have some chronic pain issues related to the reconstruction, but is now out 10 years. I just saw him in follow-up a couple of weeks ago. He remains disease-free and is doing quite well. So I'll conclude with a couple of thoughts. One is that the goal of treatment for these primary bone tumors is cure or prolonged disease-free survival. Our surgical management strives for maximum tumor removal, usually via en bloc resection for the aggressive, benign, and primary malignant bone tumors. These procedures are technically demanding, and obviously an experienced multidisciplinary team approach is necessary to provide optimal outcomes. But I want to leave you with a couple of thoughts. It's not perfect, right? These are large and fairly morbid procedures. Successful en bloc resection does not guarantee you a cure. You probably remember from the graph I showed on the mobile spine chordoma that 35 percent of patients with a perfect en bloc resection and negative margins developed recurrence by 10 years. So there may be microscopic disease that we're not resecting with our en bloc resections. And also not all tumors are amenable to en bloc resection. So we need to recognize our shortcomings and be thinking about ways to overcome this. So I think important areas of investigation that are occurring right now and are ongoing is, can we use some of the less invasive procedures, maybe some of the procedures you saw in the first couple of talks, to treat these patients? Maybe we start with cases where en bloc resection is simply not feasible. There's circumferential involvement around the spinal cord, or the neurologic morbidity is too great for the patient to accept. Maybe we try separation surgery plus radiosurgery or lit plus radiosurgery in these cases and get a sense as to whether those less invasive approaches can work in these diseases. Obviously, if your en bloc resections are not achieving 100 percent local control, you need to look at ways to enhance that result. How about radiation therapy? Can we define the optimal roles of radiation therapy? The radiation therapies we're talking about here are high-dose radiation with complications. How do we balance the complications of adding the radiation versus the potential benefits with regard to local recurrence? What type of radiation do we use? Is it protons? Is it spinal radiosurgery? Is it carbon ion? And when do we do it pre-op, post-op, both, or can we use it alone? I think these are questions that we need to evaluate. And much like what you heard in Zia's talk on the metastatic disease, we're nowhere near as far as understanding the molecular mechanisms of these primary bone tumors. They're much more rare, but progress is being made. We are beginning to understand molecular markers of prognosis that will help influence our treatment paradigms, molecular markers that will guide treatment, and we're beginning to recognize therapeutic targets against which we can use conventional therapies, some of the biologic and targeted therapies you've heard about, and even the immune checkpoint inheritors. So I think that the things that we've learned in the treatment of metastatic disease may ultimately come to bear on these rarer diseases, and I think that work is in promise. It holds promise, and the future is bright. Thank you very much.

primary bone tumors

spine

en bloc resection

chemotherapy

radiation

multidisciplinary surgical expertise