Adapted from a presentation by Dr. Oheneba Boachie-Adjei. This article is written for a healthcare professional audience but may also be of interest to patients. Please also find HSS continuing medical education on this topic at HSS eAcademy®.

When we talk about scoliosis we refer to a curvature of the spine, but within the curvature so many changes occur in the spine-especially in the adult patient-that it becomes something other than plain old scoliosis. This morning I am going to speak more about a subset of patients who not only have scoliosis, but also some abnormalities consistent with degeneration of the lumbar and lumbosacral spine.

If we look at scoliosis defined as a curvature of the spine in adults, it is prevalent in about 25% of the population. The actual incidence is probably not known; maybe about 500,000 patients have adult scoliosis in this country.

In the adolescent age group the incidence is around 2% to 3% of the population, so if you extrapolate that figure, the incidence is probably going to be about 5% to 10% in the older population. The most common etiology is idiopathic, and it occurs more frequently in females than in males and it is a carry over of the adolescent scoliosis, which is more common in girls-at least 10 to 1 compared to boys.

There are also congenital and neuromuscular conditions that can lead to scoliosis and then there is degenerative scoliosis, which is defined as scoliosis that occurs during adulthood as a result of degenerative changes of the spine affecting the facets and the intervertebral discs. These patients did not have scoliosis when they were young and they are a tough group to identify because unless you have previous x-rays or previous evaluations, you will never know whether this was an adult onset or preexisting scoliosis. Obviously there are many people with 10-degree or 15-degree curves in the lumbar or thoracic spine that never bother them and they do very well. So when they get to age 50 and they have a 10-degree lumbar scoliosis with degenerative changes and translation, is it true degenerative scoliosis or idiopathic with degenerative changes? That is a tough one. Neurofibromatosis is another condition that can cause scoliosis.

Progression has been studied by several investigators. This is the study that stands out to date. The study was done in Iowa by Ponseti and Associates. If there is severe rotation located to the right convexity and also if the L5 position is tilted, then it means that there is an oblique take off from the sacrum. That puts the trunk off balance and the patients have a tendency to progress with translation above the L5 level.

Also if the trunk is shifted, then the center of gravity is off midline so the plumb line is not through the center sacral line to the middle portion of the sacrum. Therefore there is also segmental degeneration and instability. The degenerative scoliosis exists in patients, who have previous x-rays to confirm that they didn't have any scoliosis. Post surgical lumbar lumbosacral deformity can result from a laminectomy, instability, or spondylolisthesis. This condition probably occurs in about 6% of the general population. Here is a classic posture of a patient with trunk decompensation and degenerative scoliosis whose x-rays show the tilted L5 and trunk deformity. This is not characteristic of idiopathic scoliosis.

Patients with idiopathic scoliosis tend to have balance curves. The pathogenesis of degenerative scoliosis includes asymmetrical degeneration of the intervertebral disc with overloading of the concavity of the spine. As the facet joints become overloaded, osteophytes develop with resultant translational shift of the spine and pelvic obliquity.

Kostuik et al, Spine, 1981; Jackson and Simmons, Spine Dickson et al JBJS, 1995; Hu et al, JSD, 1996

There are other patients in the elderly group (those over 66 to 75 years old), who may or may not have a history of scoliosis, and developed progressive deformity as a result of metabolic bone disease like osteoporosis or osteomalacia. Studies have shown that about 30% of patients from age 50 to 54 with a 10-degree or more curve have some deformity as a result of one of these diseases. Significant deformity makes it difficult for this patient to function and even maintain adequate nutrition. I have seen several patients who cannot eat a full meal because of the c3ding of the abdominal contents caused by the deformity. The diaphragmatic excursion is very limited so it doesn't take much for the stomach to get full and they become malnourished and lose weight like this patient. I have operated on several patients with this condition and they have really done well in terms of gaining weight, and being able to function.

Convexity of lumbar/thoracolumbar curves

In terms of patients who have increased severity and progression, symptoms can be present in patients with scoliosis who have progressed and those who haven't. Scoliosis is a descriptive term that we use to describe the spine that is rotated, curved, and translated. The process is superimposed by degenerative changes that can cause stenosis and affect the nerve roots and these changes don't go away, even beyond age 75, as you see in this patient.

Since the spine is a mobile organ structure, it does not spontaneously fuse until beyond 70 to 80 years of age. So these patients experience a prolonged period of agony. They also reach a point where they become inoperable on the basis of age and associated risk factors, including poor bone density. It is very difficult to stabilize the curvature or try to straighten it. It is just too risky. So for progressive deformities it's best to intervene early, in the 40s, 50s.

The pain that results from scoliosis is either mechanical, stenotic or radicular. One way to ascertain the type of pain involved, from the clinical point of view, is to observe how the patient sits. If this patient lets go of her arms, her trunk sinks in and at this point she has pain after 10 to 15 minutes.

This patient is seen here, applying longitudinal traction to her spine in order to relieve some of the mechanical stresses causing pain. Some patients assume this standing posture, trying to brace themselves so they can relieve some of the pressure in the spine. This is her x-ray. The question is where is the pain coming from? Patients-even those in the younger group-can point to the location of the pain on the convexity of the spine. This posture illustrates convex region pain in a 30-year-old patient. The convex pain is usually related to increased muscular activity and stress. These patients usually have to lie down more frequently and have to stretch to relieve the pain. The muscles fatigue because they are constantly working to realign the patient. That is why they are referred. Convex pain is usually muscular pain.

The same thing can apply to the older patient. If you ask "Where is your pain"? They point to the convexity of the curve. They can also get pain on the concavity. This patient has pain in two areas, the convexity and the concavity. Concave pain is usually due to the impingement of the facets. On the concavity there is overloading of the facets, resulting in arthritis type pain.

It is similar to osteoarthritis of any joint whereby joint impingement and loading causes pain. This usually occurs at the apex of lumbar curves L3-4 or L2-3 causing subluxation in these areas which can also put traction on the nerves.

Let's look at the patients who have idiopathic scoliosis versus those with degenerative scoliosis. Stenosis related to pain is more common in the degenerative group where as, mechanical pain is more common in the idiopathic group. The curve magnitudes are also much greater in the idiopathic group. So if you see a patient with a 50-degree lumbar scoliosis, it is very unlikely that it is purely degenerative.

Degenerative curves range between 20-30 degrees and usually occur in the lumbar and lumbosacral region. T12-L1 is relatively normal. So the back pain can also be discogenic from herniation of the disc. The pedicles can also get kinked in the standing position resulting in radicular pain. Pain improves with sitting or lying down. So you can examine these patients and find nothing wrong with them because of dynamic process. If they are not walking and they are not doing anything, they have no problem.

Nerve root entrapment is also part of the pain picture. These are diagnostic images showing the translation and subluxation of the vertebral bodies, and here you can see the kinking of the pedicles causing radicular pain in this particular patient.

In a myelogram, unless you do flexion extensions as well, you may not see where the problem is. It is better than an MRI in cases like this. It is in the concave lumbar region where the degenerative changes leading to the stenosis. There is the subset of patients in whom back pain results from spinal stenosis. They usually present with lumbar or lumbosacral deformity. There can also be thoracolumbar or lumbosacral deformity or as well as a double major curves. These patients can be in both categories of plain idiopathic scoliosis with superimposed degenerative changes, or degenerative scoliosis with degenerative arthritis and impingement.

This shows a typical idiopathic thoracolumbar curve with degenerative changes below that, and that will be a double major curve with degenerative changes within this area of the spine. There are even some neuromuscular patients who develop degenerative changes. I have seen polio patients who have degenerative changes, in the lumbosacral region especially if they walk and there is a lot of trunk movement.

In the evaluation we get full standing, 36 inch x-rays so we can measure all the curves and assess the extent of tilt on the lower lumbar region. We obtain lateral and bending x-rays also to get a pretty good idea about flexibility and balance. The sagittal balance in these patients is important. As the deformity progresses they tend to lean forward. The discs collapse, they get shortened anteriorly, the extensor muscles get weak, and they keep going forward. They develop a functional flat back and then they have to start bending their knees to ambulate, with resulting hamstring tightness and hip contractures. So rehab becomes very critical in the early stages. If you are contemplating a surgical procedure, you have an idea of how much correction you can achieve and how you are going to be able to balance the patient also so that you can maintain both coronal and sagittal plane balance.

We proceed with other studies like myelogram CTs and MRI scans. In some patients we do both because they complement each other. As the deformity gets very severe, it is very difficult to get all of the information on an MRI scan, myelogram CTs are helpful because we see continuity of the dye column, and flexion extension views can tell whether there are dynamic changes causing stenosis.

Also the dural sac can be better defined with contrast dye. See here, the dural sac is the same as the bony canal at this level and if you come down to where the subluxation is, you see the hypertrophic ligamentum flavum and the capsule; so the sac is reduced to less than 50% of the bony canal. If you did a plain CT you would probably not have been able to appreciate the extent of central and foraminal stenosis in this patient.

The good thing about the MRI is that you can get very good coronal cuts. It can give you a good idea of how the nerve roots are coming out, if there is an impingement or a subluxation and their extent. We use both the T1 and the T2 weighted image. The T2 weighted image is pretty much like a myelogram so you can get a good definition of the canal and also of the health of the intervertebral disc. From this information you can get a good idea as to what level is normal and where you can stop the fusion. The last thing you want to do is stop a fusion at an unstable segment, a degenerative segment, or a painful segment.

This data will also give you some information about the canal itself. I have a hard time reading some of these. Everything is gray, but the radiologist can help a great deal in terms of the findings, what is stenosis and what is not. This patient was considered to have some foraminal central stenosis.

How about discography? It is very helpful for us in the young and middle-aged adults. As you get into the older age group there is a significant amount of degeneration and it becomes less useful. Discography in the younger patient is helpful in identifying the painful levels.

Also if you are doing a fusion, you need to know where you can safely end the fusion. If the disc appears to be normal on an MRI but is dehydrated and dark, or the patient has low back pain, you have no way of telling whether that dehydrated disc is painful or not.

We know that dehydrated discs are common in the adult patient. We all have some disc degeneration, but we don't all have pain. So then it's important to ascertain whether the degenerative disc is a painful disc. Abnormal discograms and painful discograms are found in about 88% of patients with idiopathic scoliosis. These are in the young adults, 40, 50, and some 60-year-olds.

Once you get into the degenerative group, those with superimposed degenerative changes on idiopathic scoliosis or degenerative scoliosis, you get about 100% abnormal discograms. They are degenerated and inject dye goes everywhere. There is no pressure that can be contained within the disc and no pain will be elicited.

Painful discograms are more common in the idiopathic group versus the degenerative group. The test is quite useful in determining what levels to fuse to.

As far as treatment we take patients through nonoperative management first, treating the pain, with modalities such as heat, analgesics, physical therapy, stabilization exercises, and flexion exercises, depending on the type of deformity. We tailor the physical therapy to the patient's needs. If they have degenerative changes with neurogenic claudication, we don't recommend use of treadmill. The patient will do better on a bike or with swimming exercises. We advise patients to do respiratory exercises, and if applicable, to stop smoking since it helps with the bone density. If they are contemplating surgery it is advisable to stop smoking. It also helps reduce the chances of pulmonary complications.

Those that have stenosis can get significant relief with epidural injections or transforaminal injections. I have seen several patients who have avoided surgery just by using these modalities. You have to have good indications before you consider surgery, because the surgery to fuse the spine is not physiologic. Fusing the spine makes the segment immobile and stiff. Unfused segment hopefully may last a few more years before it degenerates to require a fusion.

What happens is that there is transfer of load to other unfused levels. Long fusion of the spine may affect the sacroiliac or the hip and knee, but until we can replace the entire spine, there is no better alternative to treat severe deformity than arthrodesis.

We have advised some patients to use a molded orthosis to help with mechanical pain. They use it for a short period of time and stop before they start becoming dependent on it and the muscles become deconditioned. But in some older patients I use it as indefinite treatment. An example is this 75-year-old who is not medically fit for surgery and doesn't have a significant amount of stenosis. She has responded well to epidurals over a four to five month period. She has mechanical pain and she cannot walk too far so the brace gives her support for walking and other activities. She takes it off at night. She exercises to maintain bone density and prevent further bone loss. That is a pretty good scenario.

If a patient has metabolic bone disease, it is necessary to treat that first. I refer them to the metabolic bone service, Dr. Lane's group, to get the osteoporosis treated with regimens such as Fosamax, Calcium, Calcitonin and other treatment methods that they use. In addition to having a major medical work-up, the patient is assessed for thyroid function, or renal or parathyroid problems.

But if the patient has failed conservative treatment after they have gone through all these modalities, and has deformity and stenosis, then we might consider surgery. Those with deformity with low back pain and radicular symptoms and degeneration are a tough group. The adolescent or the young adult who comes in with a 60 degree scoliosis is otherwise healthy. They may have pain, but they have no stenosis, no degeneration, no segmental instability. So we measure the curve, assess the fusion levels, and fuse them to whatever levels we chose and you are done. You don't have to worry about the junctional levels until maybe 20 years later when they come in with junctional degeneration especially if they haven't been properly aligned, functional deformity may result from excessive activity causing the unfused segments to eventually go through an early degenerative process.

If you fuse a 13-year-old to L4, 20 to 25 years later, at the most, he or she is going to have problems at L4-5 and L5-S1 levels. So I tell them to take it easy a little bit, and avoid excessive high impact, rotational sports and activities, no other things that will cause early degeneration. But if you fuse them to L1 or T12, they can do very well for the rest of their lives, provided the remaining lumbar spine is properly aligned and has not shifted.

But if we have to do surgery then we need to think about these issues. Surgical options include decompression, stabilization of the deformity, obtaining an arthrodesis, balancing the spine and restoring function. These should be the goals in this particular group of patients who have deformity with degenerative changes with stenosis. Not only do we stabilize the spine in a young adult, but we make sure that all levels that are stenotic get decompressed adequately and also restore the balance in both planes to achieve good function; and that is quite a task.

To get the best results, all of the following are important: how you select the patient, how you plan the surgery, the surgical technique that you utilize and the postoperative management. We will go through some of these issues. Surgery can be performed with posterior fusion with instrumentation, or an anterior fusion with instrumentation. You can also perform a combined anterior fusion and posterior fusion with instrumentation either front or back, or you can do a release only in the front and then follow up with posterior fusion with instrumentation. What you choose will depend on the type of patient and curve. Segmental instrumentation provides us with stability of the spine so we can get an early fusion arthrodesis, correct the deformity, balance the patient and in some cases, avoid the use of an external brace in the postoperative period.

Before spine instrumentation, everybody was placed in a cast which was not pleasant. Aposterior fusion and instrumentation works well for mild to moderate curves. The patient can achieve good sagittal and coronal balance. If you don't need to go to the sacrum with the fusion that is also beneficial. There are decompensated thoracolumbar lumbar curves for which it is necessary to fuse to the sacrum. If posterior fusion alone is done there is a very high pseudoarthrosis rate in such cases, reported to be up to about 49%. There is also higher loss of correction and incidence of imbalance. In these cases, you should consider a supplemental anterior fusion.

This is the perfect case of a 42-year-old patient with low back pain. She is well balanced. She has a mild to moderate curve. You can see that she has some early translation and degenerative changes below. If you leave this alone, within the next five years this will start sliding off to the side. So if she has demonstrated progression or has symptoms that are significant and she has failed conservative treatment, this is the best time to treat this patient. She is flexible on the bending x-rays, and the lower portion of her spine realigns very nicely. We can consider other diagnostic studies like MRI which will give information about the health and condition of the discs below.

Based on the standing x-ray, I chose L1-L2 as the end of the fusion-that is where the curve ends, that is where the stable vertebra is and when she bends over to the side you can tell that the remaining lumbar spine balances itself. The righting flex is the mechanism that allows us to stand upright so we don't develop an abnormal gait pattern. By this mechanism the body will use other parts of the spine to maintain normal balance. So if you have a flexible thoracic curve and treat the thoracolumbar/lumbar curve spontaneous correction can occur to result in normal balance.

This patient has no herniations and the central canal looks pretty good. She has a good end plate and no narrowing. So if we chose L1 or L2, it would be fine in this case. Her pain is mostly in the convex portions of the thoracic spine, so the simplest operation is posterior fusion to the L2 level. You can see how this balances out very nicely, the disc is horizontal and she has done very well for the past five years. This patient is working full time. She couldn't have had any better treatment at this stage. Waiting longer is only going to cause more problems, a bigger operation and a longer fusion. So if patients comes in with mild degeneration, it is best to treat the major deformity and allow the mildly degenerated segments to correct itself.

You can also do anterior fusion only, with instrumentation, if the patient has a flexible thoracolumbar or lumbar curves of 60 degrees, has no kyphosis within the area of the deformity, and if the remaining spine is flexible (ie. the thoracic or the lumbosacral). If the unfused segment is not flexible then they will be left with an oblique take off after you correct the primary curve. Once you talk about fusion to the sacrum in an adult then you are thinking of anterior and posterior just because of the problems of pseudoarthrosis and loss of fixation. The patient must have good bone density because if you create a long lever arm on a weak bony segment, it will collapse.

This is a patient in her 40s who came in with a very mild lumbar curve, and had a problem with curve related pain. The thoracic curve is mild. No abnormal sagittal alignment, good bending x-rays on both sides. On the MRI you can see this is a normal L2-3 disc, and below are degenerative,L4-5 discs. There is no stenosis and no radicular pain. The pain is all mechanical. We have two options, leave the lumbar curve alone and let her continue to have pain. Or, you can chose to treat the lumbar curve and hope that you can realign the remaining part of the spine and change the mechanics of the lumbosacral region and give her pain relief.

Now the difference between the two choices is that once you fuse to the sacrum and the patient's scoliosis doesn't correct (which it doesn't in most cases) or she continues to have scoliosis-type pain, then you are committed to fuse the curve too. Then you have given her a total spine fusion and that is a big commitment. So you are better off starting from the top and leaving the lumbosacral alone, because if the pain is reduced by even 50% she might decide to live with it and wait for a later arthrodesis to the pelvis.

I chose to do just an anterior selective lumbar, and the reason why this works better than the posterior is because once you remove the discs, you can completely derotate the spine in the thoracolumbar-lumbar region and provide her with almost 100% correction, which you cannot do posterior with most systems to date. By providing her with anterior column support, you establish lordosis and now the L3 to sacrum has completely horizontalized. She had complete relief of the low back.

Now we don't have to worry about arthrodesis to the sacrum. She may need it, but hopefully not until several years from now. In a 42-year old, it is much better waiting 20 years from now, than starting from below and then realizing the curve is still there and painful, and then needing to do a total spine fusion.

These are pre- and postoperative x-rays. See how much rotational correction is achieved just by the lumbar correction with the anterior instrumentation. Her pain in the convexity is gone and the low back pain is also gone. Here is another patient who is 58 years old with a much bigger curve-notice the subluxation? The sagittal alignment looks good, the thoracic curve is also quite significant but it doesn't bother her.

You can live the rest of your life with a 50 degree thoracic curve. If you are lucky, it won't cause pain. It won't cause pulmonary problems, no. The patient may be able to live with the cosmetic deformity, but it is difficult to live with progressive symptoms of subluxations and spinal stenosis. She tried other conservative measures, but was still having pain. On the bending x-rays we can see that she has some flexibility of the lumbar spine.

I decided to just do a selective lumbar like we did on the previous patient. See how the thoracic improved spontaneously to 32 degrees? And we maintained her sagittal alignment. Even though this is a single curve, it is a bigger curve and I think I was pushing the envelope. She broke her rod and I had to go back and place a posterior short-segment construct on it, which took care of the problem.

That was anterior fusion which is useful for flexible curves with minimal kyphosis, fusion above the sacrum, and good bone density. But with the anterior procedure that you cannot decompress and stabilize the painful or stenotic segments. This technique gives you indirect decompression,by restoring alignment you are opening the concavity of the fractional curve where there may be mild stenosis or impingement.

But if there is fixed stenosis from osteophytes, herniated disc or an ankylosed segment, this curve correction won't do it. You have to approach it posteriorly to be able to achieve that and that is where the anterior release and fusion alone comes in, and then we will place the instrumentation in the back and perform decompression also. For arthrodesis to the sacrum, a circumferential arthrodesis will increase the fusion rate and provides correction. Posterior approach can also allow the performance of the decompression.

Although other surgeons around the country still do this, there is no need to put both anterior and posterior instrumentation. Once you put instrumentation in a segment of the spine, the correction that it provides is final and another instrumentation in the back will not give you an iota of correction. It might help as additional stabilization if the anterior instrumentation is not strong enough.

We did a review of 28 adults fused above the sacrum. Radiculopathy was present in 7%, posterior fusion was performed in 8%. Most of the patients had combined procedures and good to satisfactory outcomes. The scoliosis correction was 61%. The patient satisfaction was 71%. There was no pseudoarthrosis. There were complications, including junctional deformity and infection, in one patient each.

If the patient needs a fusion to the sacrum and you choose them appropriately, and you select the proper technique and approach, you can get very satisfactory results.

Now what about the patient who has lumbosacral lumbar deformity or thoracolumbar lumbosacral or double major curves? I looked at patients in my practice who had this type of deformity. Their average age was 66 and there were more females than males. A lot of them were idiopathics who have now developed degenerative changes. A combined procedure was done in 15 of the 16 and this was done in one-stage. Here is the clinical and radiographic outcome in all of the patients. This is the average blood loss for the anterior and posterior, operating time for both and then the number of levels that were fused, 7 in the anterior and 13 in the posterior.

The listhesis that was mostly present at the L3-4, as you have seen in the x-rays that I have shown. L3-4 appears to be quite significant for translation. The lumbar/ thoracolumbar curve ends at L3 or L2 and the spine is rotated clockwise and then the other segment is rotated the opposite direction. So you already have a twist. This is potentially a very unstable segment which results in translational shift of L3 or L4.

Stenosis was also present at the same levels, so where there is this high activity of movement and instability, there is also bony reaction and hypertrophy of the ligamentum flavum and the capsule. Stenosis quite often occurs in L4-5 and S-1. Direct decompression was done in 10 of the 16 patients. Six patients had decompression by indirect corrective maneuvers. That is why if you reviewed the MRI, the CT scan, you may have guessed foraminal stenosis on the concavity, but there is no osteophyte, no disc herniation, and there is no ligamentum flavum hypertrophy, there is just narrow space. Curve correction can open the space in these areas and we don't have to go after the nerve root foramina.

Major scoliosis correction averages 53%. Lumbosacral fractional curve correction was 46% in this group. The key really is balanced correction. Kyphosis approached the physiological range also, and the same thing for lumbar lordosis. Some of these patients actually came in with lumbar kyphosis as much as 30 degrees so they were really out of balance. They were corrected to within normal physiology lordosis and spinal alignment.

Balance in the front was improved from 4.5 to 1.8 cm, so they were shifted to the left and the right by that much, corrected to the average physiological range. With regard to sagittal alignment, they were leaning forward by an average of -9.4 cm, and they all came within the physiological range also. This is an example of the patient with the lumbar deformity and thoracolumbar kyphosis, so the trunk is really short in this patient. Segmental translation is present on the bending xrays and then the 4-5 and 5-1 disc spaces look pretty good. We did a combined anterior and posterior fusion to L4. This is indirect decompression. We didn't have to do direct posterior decompression and her symptoms improved. Her short trunk significantly improved. This is a picture of the transverse abdominal crease which is much better after surgery.

Here is another patient at 52 with progressive scoliosis, low back pain, and L5 radiculopathy. You can appreciate some of the osteophyte in this area. This is an MRI scan, so we chose the direct decompression by the discectomy anteriorly and then posterior we did a limited foraminotomy. If you do a decompression, outside your fusion range, you want to be very, very conservative. If there is thick ligamentum flavum, do a foraminotomy, don't take the disc, and don't remove the facet joint. The moment you take the disc or the joint you have created an unstable segment and if your fusion does not include it, it will eventually collapse.

So you can do a foraminotomy for the stenosis, below the fusion, without fusing it. In her case, this was a problem so we just took the ligament flavum and did a foraminotomy and tried to decompress it and then stabilize her to L4. The foraminotomy was done below and we didn't have to go to the L5 level. She did well, for 6 years and now may be facing extension to the lower spine.

Here is another patient with low back pain and L5 radiculopathy. You can see the translation at the L4-5 level. Here we are dealing with direct compression. She has good bending x-rays on the front and back. You can see the changes in the end plate and these are not enough cuts. There was some foraminal stenosis also. Here we did an indirect decompression with instrumentation anterior and posterior. We fused her to L5 because we could not fuse her to L4 owing to the extent of the degeneration. Even if you can decompress indirectly, if the L4-5 is so degenerated, you are just going to be asking for more problems if you fuse above that level.

The long fusion to L5 patients don't do very well, so I am very very careful in selecting them. I have probably done five L5 fusions in adults. This patient is still doing very well; a couple of others are still hanging in there, but what happens is, it just shifts all the load to the 5-1 so it doesn't take long for 5-1 to degenerate and become unstable. So in most cases we extend the fusion to the L5-S1 level, to avoid refusion and extension.

If that happens, you can approach the disc from the back with interbody fusion. But, if there are changes we don't hesitate, we just go straight to the L5 level. This is a front and back rotation on the bending x-ray. So even when you fuse to L5, they can still bend forward, most of the forward bending is hip movement. They may not be able to sit in the bathtub but if they have well-stretched hamstrings, they can do very well. So postoperative physical therapy becomes very important. Even pre-op get them to stretch the hamstrings and maintain good flexibility so that they can achieve this kind of motion after surgery.

Patients are going to ask you, "Can I bend forward or sideways?" If you have to go to the sacrum, then you have to do an anterior fusion. The anterior fusion also has to be supported with structural grafts, because most of the load is not going to be seen in the front, once you remove the disc you have created a void, and if you just use cancellous bone, it is not going to be adequate to support the anterior column. So either use bone or titanium mesh grafts with bone to provide anterior column support.

These are the iliac vessels, vein, common iliac, and then the external and internal iliacs. So we go, between the bifurcation and then take the disc at L5-S1 and put a bone graft in there, and do a posterior instrumentation.

This gentleman came in at age 70. He has this deformity which looks like degenerative scoliosis. He has degenerative changes and a significant curve. He also developed weakness of the extensor muscles which contributes to the postural deformity. His thoracic spine is not that bad. He has a severe trunk malalignment to the front. With the bending x-rays and hyperextension films you can see how much we can get him back if we just fused it posteriorly within the lumbar spine. You can also see the osteophytes anteriorly and the myelogram/CT shows the stenosis at 4-5 level,with the osteophytes. There is also hypertrophic ligamentum flavum and also, a very tight spinal canal.

You cannot decompress this indirectly with just spinal column realignment. We had to do thorough decompressions. We went anteriorly, did a release and placed structural grafts and then performed posterior direct decompression and pedicle fixation to the thoracolumbar spine, and Galveston fixation to the pelvis. When you create such a stable foundation you can establish a long lever arm above it.

By correcting the lumbar spine and establishing lordosis, his thoracic spine takes on the natural kyphosis. It happens naturally and that is how you treat this kind of patient. That is before and after photos. His sagittal balance and coronal balance were maintained.

When you have a trunk translation like this, most of the stenosis is dynamic. It may not be a direct compression from an osteophyte or a nerve root, but this patient has degenerative scoliosis. See, she has translations here, and osteophytes there and her problem is really one-sided. You can see the big osteophyte projecting into the canal, so this has to be removed. You cannot just straighten the spine and leave this alone.

So here we do a combined anterior fusion to the sacrum. This technique uses the complete release of the spine and makes it more flexible. Then, posteriorly, we do a thorough decompression and instrument, so we can shift her trunk back to the midline. We created a long lever arm in order to align her trunk over a level pelvis with Galveston fixation. Anterior column support transfixed with screws through the bone grafts was performed. That is the follow-up x-ray.

Q. When you can reduce the patient before, you do an anterior procedure first right?

Q. So how do you control spinal stability.

A. That is a good question. Instability doesn't happen with anterior release. As long as you have your posterior elements in fact, the spine is still stable. We can even get these patients to get up and walk in between stages. Removing the disc and replacing it with bone graft does not make them unstable at all. If you leave the patient alone with no further surgery, then there could be more progressive deformity, but they will maintain their neurological function so long as the posterior elements are intact. If you do a big resection anteriorly then the spine can become unstable because you have created a huge void.

This is a patient that we treated who experienced one of the complications that you can see: painful iliac screws that were removed, infection, hook dislodgment with junctional kyphosis, and breakage of the rod. Two patients in our group (12%) had pseudoarthrosis, One patient who we fused to L5, developed a collapse because of weak bone, so we had to extend the fusion to the sacrum.

This is an example of a 58-year old patient with severe deformity to the sacrum. You can see the osteophytes. This patient has spinal stenosis at this level in addition to severe scoliosis. Bending x-rays shows this doesn't move at all, and her myelogram CT shows a very tight spinal canal confirmed also on the MRI scan. This is where I like the CT myelogram better than the MRI even though some people claim that they are pretty much the same in terms of findings. I prefer the myelo/CT.

With the combined procedures, anterior and posterior, with osteotomies and decompression, we thought we had realigned her spine, but she is still 8 cm off and the reason is that her lumbar curve was over corrected. The amount of correction is not the critical issue. You want to balance the patient. If the patient is balanced then that is good. But this patient was off balance.

Now about leaving the patient out of balance-you can see here that it doesn't go away. Her decompression has relieved her stenotic symptoms completely, but she stays out of balance and continues to stay out of balance and when she comes back she is actually worse. If that happens, either a rod is broken or you didn't get a fusion. Without a solid fusion the patient is depending a lot on your instrumentation. So putting all that stress on the implants, she actually broke the rod. We had to go back and perform an osteotomy through the pedicle unilaterally. That is the nerve root. We then created a wedge underneath the spinal column and then closed it with implants in order to balance her spine. That is interesting, to correct imbalance with additional curvature.

We increased her scoliosis and now we can put her head over her pelvis. This is the structural graft at the level of the osteotomy. She has remained balanced and she is doing quite well. This is before, when she was balanced, and then when we put her out of balance, and finally we got her back where she is.

So the amount of correction is important and any time we see a patient out of balance, we either warn them that we may have to go back and do something or fit them with a shoe lift. We don't rush everybody right back to surgery. Because some of them do well, and fuse very fast. The bone takes over so the hardware can rest and some patients also do well with a 1 to 1.5 cm shoe lift.

We graded this group of 16 patients, in terms of their response, as excellent, good, fair, or poor. The majority of the patients were very satisfied. They were able to get back to exercise. Pain relief was significant.

There is another method that I have utilized in a select group of patients. It is an attempt to make the operation simpler and less extensive and to preserve motion segments. Like this particular patient. The thoracic deformity is not that significant. Her problem is really lumbar and lumbosacral and she has good bending x-rays. Her MRI scan showed this translation and the instability at the 3-4 level. She had degenerative levels throughout the lumbar spine.

Here we are going to fuse anterior to the sacrum and posterior to the sacrum. We realized that most of the posterior instrumentation has been very long and that is really the only way that you can get the patient balanced in all the planes. But we thought that if we did just anterior release and corrected the deformity anteriorly to a certain degree, then it would make our posterior instrumentation short.

So here we do the anterior release and we release all the segments to the sacrum, but we instrument the main curve with two rods as opposed to just one rod, and we put structural grafts in between. We can correct the primary lumbar deformity. Now when we turn the patient over all we have to do is the distal decompression and fusion. So this is like a combined degenerative and deformity approach and we leave the thoracic alone.

The patient has a very small scar on the back. This is not different from the anterior fusion only surgery anyway, because the scar is the same, dissection is the same, and the only difference is that we have corrected the spine with instrumentation from the front. So when we go into the back we don't have to put instrumentation in, and we actually are getting better results with this technique. You won't find it in any textbook. I have been doing this for the past four years and hopefully we can get it published.

Overall, the results are very good. Most patients are happy. They can go back to work and other activities. They can exercise and there is a very high subjective patient satisfaction.

So in all patients with scoliosis, lower back pain, and stenosis, we should be aware of the full scope of the problem, the functional limitations, and the radicular signs. Then we can institute the appropriate treatment. They are mostly thoracolumbar-lumbar curves with fractional lumbosacral curves. Their pain can be very disabling. It may be mechanical, radicular, stenotic or a combination of all three. It may be treated conservatively with exercise therapy, injections, metabolic bone work-up and treatment, and when they fail then surgery is necessary.

The hybrid, or the combined procedure provides patients with limited fusion. It preserves the thoracic motion segments, increasing the fusion range because we have the interbody to fuse, and, as you saw in the last x-ray, I didn't go to the pelvis. That is a plus because if you can avoid putting fixation into the pelvis then you free up the sacroiliac joint also. This goes against orthopaedic principles, to close the joint and not fuse it, and most of these patients come back with pain in the area of the iliac screw so we have to remove it.

We also get better correction anteriorly and hopefully we can prevent junctional deformities. Thank you for your attention and I will be glad to answer any questions.

• Imaging of Lower Back Pain by Leon D. Ryback, MD

Q. That was a great talk. Some of the patients that we see have a lot of back pain. There is also a subset of patients with old degenerative scoliosis in the concavity of the curve, who have a lot of L3 radicular pain. Typically, the group of patients we see here get three to six months relief with transforaminal injections and it gets harder and harder to achieve because the space gets narrower with time. In that patient population that just has unilateral one level leg pain but not much back pain, is there any less invasive type of surgery?

A. I have seen patients get foraminotomies for that. If it is a soft disc or hypertrophic ligamentum flavum with capsular hypertrophy, you can do a very limited foraminotomy to relieve the symptoms, but you have to warn the patient that it may be temporary like your epidural or transforaminal. Give him six months to a year and then they will be back because that segment is weak. It is concave and it is going to collapse. So now apart from the capsular hypertrophy and ligamentum flavum, then the pedicles take over and when that happens, you can't do anything. You can't remove the pedicle, you can't shave it down, and you have to open the Space, with direct decompression fusion and instrumentation.

Now in the younger patient, if they have a mild curve, you can do a limited arthrodesis, but it is difficult to do without worsening the curve. The stenosis usually happens to be at the end of the curve. Now the moment you do that, you are transferring the distress to the apex of the curve and that is the worst place to transfer stress.

Q. In the degenerative scoliotic patients we felt the source of the pain was from the stenosis itself. So when you say pain from stenosis, do you think there is a particular pain generator. What you would identify as the pain generator in patients who have stenotic pain?

A. I think it is neurogenic claudication with possible compartment type syndrome. There have been several theories proposed to explain why patients have neurogenic claudication when they are walking. Is it from a compartment type syndrome with poor vascularization or poor venous drainage? But it is usually neurogenic claudication and those patients have a very narrowed central canal. If they have foraminal stenosis they get radiculopathy. So it is really within the segments that are translated or severely degenerated that have developed actual mechanical stenosis.

Q. I have heard some theories about axial back pain, from the vascular component.

A. I think I see what is a combination of discogenic or facet or muscular pain. The muscular ones you can tell it usually on the convexity of the curve. They can tell you where it is, they get tired, fatigued and when, they lie down it goes away. The mechanical pain from the facet region is usually on the concave side or the junction of the two curves. It is painful on extension and rotation. The discogenic ones have pain when they sit and flex forward. They may have good disc height but it is bad. So they are still loading it and then it becomes painful when they flex forward or they sit up for too long. So I think those are the three types of pain. Those are the patients in whom it is hard to determine whether it is true discogenic or facet pain. What happens in most of these patients is that they have pain and the pain goes away and the reason why the pain goes away is because they develop osteophytes and develop ankylosis of the facets. So they stabilize it. There is no more motion. The concavity is all bridged and most of the time we really have to osteotomize the concave portion of the curve during surgery.

Q. Joe was saying when we see a subset of patients who have spinal stenosis who present with neurogenic claudication and back pain. I think when we know it is coming from the canal because it responds to epidural. I think what Joe was trying to say was that if you have foraminal stenosis you do get nerve root compression. And unilateral pain. And still that subset of patients that have central canal stenosis and they may have the back pain with standing and walking around.

A. Those are probably patients with proximal stenosis at L1-L2 levels; they get the back pain more than in the lower extremity, especially thigh, and leg. I think the characteristic presentation of these patients is with posture also. If the back pain is worse when they are standing or walking, then I think that is part of the neurogenic picture. Especially in the buttock region. If you ask them carefully, they will tell you. If they really point to the lumbar spine as the painful area when they are standing, I am not sure if that is true claudication.

Q. Because in managing patients conservatively whether we are using physical therapy or some type of injection or other technique, we are always trying to identify it, as you are, specifically where the pain generator is and I think we often simplify say its probably facet or discogenic and missing other components if you try to manage them that way.

A. If you are doing a stress test on these patients and have them walk or bike, do you do any of these tests to better define whether this is neurogenic or mechanical? Do you do that?

Q. I question the validity of any of those test now.

A. You only go by history then.

Q. Is there bracing at all?

A. Yes I talked about bracing earlier on for acute pain and for the older patient who you can treat with injections periodically. We will brace them from a mechanical point of view.

Q. I just want to make a comment. You know we try our best to not operate on these patients and get them better and they don't have any sustained improvement. We showed a number of patients over the years where I was initially somewhat skeptical about these procedures and they have done surprisingly well a number of times. I think that probably is more your expertise. Is that not what we would see outside HSS?

A. Patient selection is very critical too. We should get the fellows to come up with a protocol so we can do some outcome studies too. That would be helpful.