Thanks, Justin. Thanks for including me and Bob as well. Thanks for inviting me back. How many of you have done AIS cases as part of your residency and involved? So, all right, not too bad. As Justin said, it's not something that traditionally we see a lot of as neurosurgeons, which is a little bit of a shame because they are nice cases, but also because the principles, a lot of the principles of deformity surgery come from the AIS patients and the work done going back historically. And AIS, in a lot of ways, is the archetypal deformity. So, understanding some of the principles, picking levels, treating these patients, I think is useful in the adult world as well. So, I'm going to run through sort of my algorithm. It's a very general imperfect algorithm, but maybe to help give those of you who are less familiar with AIS a bit of an understanding, maybe hit on some point of value to everybody. So, very brief, very simple management algorithm for AIS, which is really, it's one of the few areas in neurosurgery where we're treating the x-ray almost as much as the patient. So, first is to make the diagnosis. I also have a talk tomorrow where I'll cover a little bit more of the conceptual and background information, sort of. So, some of this I'll skip over fairly quickly. But making the diagnosis, assessing the risk of progression, there are many ways to do that. I'll talk about a couple. Determine the need for treatment, and then choosing the treatment method, and we'll spend some time, hopefully maybe run through a couple of quick cases at the end before we go to lab. So, the diagnosis algorithm that I put together fairly quickly looks something like this. So, the question is, is there a coronal curve greater than 10 degrees? And if there's not, that's scoliosis. That's something different than scoliosis, and that's the end of the discussion. Those are the patients who get screened and sent to you for evaluation. Obviously, we aren't treating 10-degree curves, but that's the cutoff for scoliosis. Patients with AIS are between 10 and 18 years of age. If they're outside of that, then you're either dealing with a juvenile. Much of the treatment is very similar for those patients, although there can be some considerations about intraspinal anomalies that are more applicable to that group than to the AIS patients. And if they're over 18, then it's adult idiopathic scoliosis. Again, sometimes we treat them similarly, but there can be some differences, particularly if they're getting on beyond their 20s and 30s. And then we get down to the area where it's important, especially as neurosurgeons, to differentiate the AIS patient from the non-AIS deformity. For example, if it's a left thoracic curve or a right thoracolumbar-lumbar curve, as we'll talk about tomorrow, that's against the typical pattern of an AIS patient. Those patients ought to be worked up further with MRI or at least further investigation to make sure that it's truly an idiopathic curve. Not all boys, this was meant to be a dashed line here, not all boys with presumed AIS need to be imaged, right? But only about 10 percent or so, 10 or 11 percent of AIS patients are male. So you need to have a little bit of a higher index of suspicion for a boy with scoliosis that there might be a cause for it rather than just an idiopathic curve. Is it an angular or atypical curve or is there a syndrome present, Marfan's or NF, that makes it a non-idiopathic curve? Are there signs of a neural tube defect? And this is really where we need to bring our expertise. Is there a hairy patch indicating a diastem? Is there a leg asymmetry, meaning leg length, calf circumference, foot asymmetry? Is there a dimple above the gluteal cleft? All of these things that should lead to an evaluation of the spinal canal. And if you exclude these, I think I got the major ones here, then we are talking about, most likely talking about adolescent idiopathic scoliosis. And then the next step is to assess the risk of progression. And here it's really an interaction between the size of the curve and the potential for growth of that child. These curves tend to progress as the child grows, with the main period of risk being that adolescent growth spurt. So these data are from a paper that I think Dan mentioned in his talk, the Lonstein-Carlson paper from 1984. But some factors that can differentiate the low from the moderate from the high-risk patient for progression. So smaller curves in a more skeletally advanced patient. Risks are grade 2 to 5. I think, again, tomorrow I have the slides showing sort of the risker grade, but this is the progression of the ossification of the iliac apophysis. So this is a little bit higher grade, grade 5 being skeletally mature. Small curve, skeletally mature, lower risk. The highest risk, of course, is going to be the bigger curves and the less mature. We also ask questions about menstrual histories in girls, family histories, growth spurt. We chart the growth of these patients, trying to assess the potential for progression in that child. There is something, a genetic test that I'm not sure exactly where the pendulum has swung on this. The Scoley score was being used pretty widespread, at least at our institution. We've backed off a little bit, finding that A, the expense isn't covered by insurance companies, and B, it may not be much better than these criteria in terms of differentiating low from moderate from high-risk patients. So you're going to tend to treat the higher-risk patient than the lower-risk patient for progression, of course, but then there are other considerations for treatment, and these are all relative. These are not absolute. With every patient, there's a conversation, and particularly with the parents as well. So if you have identified a patient who has a real AIS and is meeting these thresholds, really sort of over 20, 25 degrees, and we start to talk about management of some sort, we divide this then into operative and non-operative treatment. And so non-operative treatment, at least in the Northeast, cob angles less than 35, 40 degrees generally we treat non-operatively. I think that that threshold is different in different places for a variety of reasons, but we rarely treat anything less than 40, 45 degrees in New York. The scatterly, the immature patient, the patient who we can sort of guide their growth. So the more immature the patient is, with a reasonable curve, we're going to try to brace those patients in general. Body habitus is a factor, of course. And then bracing is going to be better and better tolerating of a better outcome for the patient as an acceptable clinical deformity compared to the patient who presents with a main complaint of the clinical deformity that we can really only correct with operative treatment. Bracing is really to hold the curve, not to correct it. Operative treatment is the way to correct a clinical deformity as well as the radiographic deformity. So in surgical treatment, the keys of moving ahead then in surgical treatment, we need to determine which curves are we going to instrument and fuse. And the principle here is to correct and fuse the structural curves while not fusing the flexible, non-structural, or compensatory curves. And then as part of that, we need to determine the UIV and the LIV. What's our construct going to be? Where are we going to instrument from and to? This first part is really where the Lenke classification comes in. And that rests first on a thorough radiographic evaluation, as we've talked about. Certainly long cassette films, upright AP or PA films, and an upright lateral where we see from ideally base of the occiput to the hips. But with kids, particularly when we're planning surgery, we then go ahead and get some flexibility films. We get right benders and left benders, and that's to look at the flexibility of the various curves. And then generally we also get another flexibility film, which I didn't show here, of some sort. I use a push-prone film, lie the patient down, put on a lead glove and a lead apron, and push on the apex of the curve while stabilizing the shoulder or the hip to see what's the maximal extent of active correction I can get for that patient. And that just gives the best sense, I think, in combination, of course, with the benders, of how flexible that curve is and how much we may be able to correct the curve. We've talked about some of this already, but for the Lenke classification, the apex, we need to be able to identify the apex of the curve. Usually we can identify that as a maximally displaced or horizontal segment. Cob levels are the levels that are tilted maximally into the curve. Sometimes this gives people starting out looking at deformity a little bit of trouble. Sometimes it's easiest just to take your goniometer and start above where the curve is. If you're trying to measure the cob angles here, start with your goniometer up above, find the first level where you're sort of maximally tilted into that curve, and follow that through. And so in this case, right, this level here, T10, is bending up into the next curve, so T11 is the proximal cob level, and then L3 is the distal cob level on this example. And it's important when you're tracking progression to measure the same levels from one film to the next. The neutral vertebra is the non-rotated segment. Look at the pedicle symmetry to identify that. And the stable vertebra is the one that's closest to being bisected by the central sacral vertical line. So we put those pieces together to then classify a curve based on the Lenke system, which is shown here. And sometimes this is a little bit daunting, I think, starting out. I know it was for me in my fellowship. So let me just very quickly walk through sort of the order that I attack this. The first thing to do is identify, right, which curves are which based on the SRS definitions of the location of the apex. So thoracic, thoracolumbar, lumbar, and then proximal thoracic up above the T2. The structural criteria for the minor curve. So the major curve, the biggest curve, is always structural. So there's no issue there. So then it's a matter of determining which of the minor curves, the smaller magnitude curves, are structural and need to be included in the instrumented construct. And here are the criteria. Again, it's worth reading this and having this. But basically, if these bend out on the coronal films to less than 25 degrees, they're non-structural. So anything that doesn't bend out to less than 25 degrees is structural. Or a kyphosis of greater than 20 degrees on the lateral film is a structural curve. That was really the contribution, one of the main contributions of this system, which wasn't just the work of Larry, but several other investigators, was adding in the sagittal plane considerations, which were not part of the King system. So with that, identifying the structural curves and the various curves based on their apices, we then can type the scoliosis. And this is simply, these six subtypes are simply based on which curves are structural and which are non-structural. And then they all have a name that relates back to the, again, back to the sort of the King system, main thoracic, double thoracic, double major, et cetera, but trying to increase the inter-observer reliability of the study by having these structural criteria. Finally, there are two modifiers. One is the lumbar apical modifier, and that's looking at where the CSVL falls relative to the apex of the lumbar curve. There's an A, B, and C based on whether that CSVL is between the pedicles, at the pedicles, or lateral body, or completely medial to the lumbar curve apex. And then there's a sagittal profile modifier based on the kyphosis of T5 to T12. So how do you pick levels? Well, first you pick curves. The major and the structural minor curves are included in the construct in general. Again, beware the rule maker, that these are some pretty consistent generalities. And we exclude the non-structural minor curves. We're trying to save levels. And so therefore, the classification helps us determine which curves should be included, but it doesn't tell us specific levels, proximally and distally, or the extent of a curve that's to be instrumented. And this is where judgment comes in. So looking at this curve quickly, so here, or this scoliosis quickly, here we have a lumbar curve, thoracolumbar lumbar, and a thoracic curve. The proximal thoracic curve really here doesn't warrant too much in the way of considerations. We can identify the apices, the magnitudes. This is 36. This is 43. So our major structural curve. So now the question is, is this, is the main thoracic curve, is that structural? We look at the right bender for this film, right? That's the one that should straighten that out, and it does. It's below 25 degrees. It bends out to 18. So that's a non-structural minor compensatory curve, and doesn't need to be included in a construct. And so we put this together. This is a 5C, okay? It's 5 because it's simply a thoracolumbar lumbar curve with a non-structural main thoracic curve. Basically all 5s are going to be type C. You can think about why that is relative to the CSVL. And then we look at the sagittal plane, and there's 18 degrees of kyphosis, so that's normal. So this is a 5CN. Okay? But it's all about balance with levels, balancing cob levels, or balancing, avoiding progression, or adding on the curve versus preserving motion segments. If you just wanted to avoid progression, or adding on segments to the curve post-operatively, I'm talking about, you could fuse at least the cob levels of the structural curves, maybe a little bit more. On the other hand, if you wanted to preserve motion segments, you would err on the side of selective fusion, fusing as few of the curves as possible. But these are the considerations, among others, that we try to balance in level selections. Other things that we look at, we look at the shoulder balance. That's a topic in and of itself, how the shoulder tilt manifests itself, and also given the flexibility of a curve and how much we may be able to correct a curve, what that's going to do to the shoulder balance. And finally, based on the flexibility of those bending films and the push prone, what the post-correction position of the LIV is going to be, how well we're going to be able to center that lowest instrumented vertebrae. You don't want to fuse to a level that's way off to the side, necessarily, or has significant tilt or residual rotation. That puts the subjacent level at significant risk for degeneration and adding on. So a real quick run-through, confirming the diagnosis, assessing the risk of progression, determining the need for treatment, and selecting the treatment. I pulled a couple of cases. Let's walk through two. We can do that in probably about three minutes, sort of how this looks in real time. So a 14-year-old, 46-degree proximal thoracic, 81-degree thoracic and a 30-degree lumbar curve. If there's any doubt about why this patient might need to be treated, I think this probably puts that to rest, a significant trunk shifts, waistline asymmetry, and posterior rib hump. Here's the radiographic evaluation. So we look at the left bender to see how the left curves bend out and the proximal thoracic. Is that structural or non-structural? So that's structural. We look distally. We're going to skip this for the moment. We look distally. This thoracolumbar lumbar curve, structural or non-structural? So that's non-structural. This is our major curve. So is that structural or non-structural? So that's going to be structural. So we have a structural PT, a structural MT curve. So that's going to be a double thoracic, right, type 2. There's 20 degrees of kyphosis. So that's normal. And the apex of the lumbar curve is basically the CSVL is between the pedicle so it's a 2AN. So we're going to fuse the proximal thoracic and the main thoracic curve here. And then it's a matter of picking levels. But basically we look at trying to get a level that's distally, neutral and stable if possible, in this case L1. Proximally we need to get shoulder control. Again, that's a little bit of maybe a more advanced topic looking at the fact that the shoulders are level and how we have to control this. But getting up to T2 or T3 to maintain shoulder control and address the proximal thoracic curve. And that's how we would walk through that. And there's the post-op films. Last one for illustration. So here are the measurements 40, 87 and 84. So we're really going to look at particularly this proximal thoracic curve for flexibility. It bends down to 37. Distally, no surprise, that stays above 25. So these are all structural curves. We also have a thoracolumbar kyphosis. And I believe, I can't see it from here, I don't have my glasses on, but a positive kyphosis. So it's a 4C, triple major. And so all of these have to be instrumented. And it's a matter of then experience deciding can we stop at 4 at the Cobb level? Can we stop a little shorter? Do we need to go longer? How many levels can we save distally? But instrumenting all three structural curves in this case. Let me end there. Any questions?

Adolescent Idiopathic Scoliosis

diagnosis

treatment

deformity surgery

AIS patients