Lumbosacral Fixation Techniques and Indications
Indications for a fusion to the sacrum in primary patients include lumbar curves with substantial disc degeneration at L4/L5 and L5/Sl. Indications for long fusion to the sacrum in patients who have had previous surgeries include severe disc degeneration at all levels below the previous fusion. This is often associated with an element of coronal and/or sagittal imbalance.
The paper by McCord et al has shown that the most stable way of fixing a lumbosacral joint is to include fixation of the ilium bilaterally that extends well anterior to the axis of the rotation of the lumbosacral disc. Saer and Winter have shown reasonable results with long fusions to the sacrum where the sacral pelvic fixation was Galveston or Galveston-like fixation combined with an anterior fusion of all lumbar segments.
The minimum requirements to achieve a long fusion to the sacrum include the following: Ideally, one should have four-point fixation of the sacrum and pelvis. This includes two bicortical S1 screws and two long iliac screws. The iliac screws should measure at least 60 mm in length. Also important is structural grafting at L4/L5 and L5/Sl to take stress off the lumbosacral instrumentation. An anterior fusion is ideally performed from T10 to the sacrum. Segmental fixation should include each level without any gaps at L3, L4 and L5. As demonstrated by Dekutoski and Transfeldt, ideally the sagittal alignment line should fall through or behind the posterior aspect of the lumbosacral disc. Failure to achieve all of these goals will likely result in either pseudarthrosis or partial loss of fixation at the sacropelvic junction which often results in a fixed sagittal imbalance/kyphosis problem.
There are several issues involved with deciding whether to stop at L5 or the sacrum. Many investigators have stated that if L5 is deep-seated, that significant breakdown at L5/Sl is not likely to occur. "Deep-seated" has been defined as big transverse processes and the L5 segment sitting below the intercristal line. However, I am not aware of any peer-reviewed manuscript that has ever tested this hypothesis. Certainly if there is a spondylolisthesis at L5/Sl, a previous laminectomy at L5/Sl, or suspicion of foraminal stenosis at that segment, then fusion should be extended to the sacrum.
Indications for posterior-only surgery in patients with degenerative scoliosis include the following. One category of patients are those with de novo scoliosis who principally have spinal stenosis. In this case, the principal objective is decompression and relief of leg pain. Fusion is performed from the lower thoracic spine or upper lumbar spine to L5 and not to the sacrum. The patient otherwise has satisfactory coronal and sagittal alignment without any thoracolumbar kyphosis or thoracic kyphosis. Once again, to emphasize the principal objective is to prevent further progression of the deformity rather than to correct the deformity. The main reason for the surgery here is to relieve leg pain. The main reason the fusion is being done is that there is a major concern that the deformity will progress after the decompressions. Usually in this kind of situation, the decompressions are lateral recess decompressions and not terribly wide ones.
The other indication for a posterior-only surgery would be a patient who is an idiopathic with superimposed degenerative changes. Most often this will be a double major curve pattern which fusion potentially can be stopped at L4 or L5. For such a patient, his/her sagittal and coronal balance should be OK. L5/Sl should be without any central or lateral recess stenosis at that level or any posterior column deficiency. Posterior column deficiency includes previous laminectomy at L5/Sl or a bilateral lysis at that level.
If fusing a young patient (adolesceint or young adult) to L3 or L4, I would recommend preserving the distal third of the ilium bilaterally. It is conceivable that such a patient will require extension of the fusion to the sacrum at a later date. If the distal ilium is preserved, then reasonable fixation of the sacropelvic unit can be achieved. If the distal ilium is not preserved, then adequate fixation of the sacrum might be questionable.
One of the significant problems with doing a long fusion to the sacrum is having an adequate supply of bone graft. In this situation, usually the lumbar spine is in kyphosis and is being converted to lordosis with the surgery. This opening up of the disc spaces requires filling the disc spaces with bone graft. Also if the distal ilium is being instrumented, then the distal third of the ilium cannot be used for bone grafting on either side. Sources of bone graft include autogenous ribs, the anterior ilium, the posterior ilium, allograft, demineralized bone matrix, and what's the future of BMP? I believe most data on demineralized bone matrix suggests that it is a reasonable graft extender, but not a reasonable graft replacement substance. In order to increase the supply of autogenous bone graft, it is helpful to harvest the tenth rib all the way back to its vertebral body. It is best to harvest the tenth rib from the outside back to the spinal extensor muscles and then from the inside from the spinal extensor muscles to the articulation with the vertebral body. This makes closure easier and preserves the spinal extensor muscles. The novice surgeon will say that it is very hard to get to the T10 vertebral body through a tenth rib thoracotomy and that taking the eighth or ninth rib makes it easier. However, taking the eighth or ninth rib usually involves disrupting more of the serratus anterior muscle. Getting to the superior order of the tenth vertebral body is difficult if one only harvests the tenth rib back to the spinal extensor muscles. But if the tenth rib is disarticulated from the vertebral body, then access to the superior portion of T10 is quite straightforward.
Also, the novice surgeon will often say that when doing a thoracoabdominal approach at times it is very difficult to get adequate exposure at L5/Sl. In a minority of patients, L5/Sl can be easily exposed behind the great vessels after harvesting the ilio-lumbar vein. More frequently, this does not give good exposure of L5/Sl. Adequate exposure of L5/Sl is achieved by extending the skin incision and exposure through the abdominal muscles somewhat more distally and working in between the bifurcation of the great vessels, the same as one would for a more limited paramedian approach.
Options of structural grafting anteriorly include cages, structural autograft and structural allograft. Structural autograft probably incorporates the fastest but is associated with fairly high morbidity. Buttermann et al recently published a comparison of autograft versus allograft in the lumbar spine in adult scoliosis patients. They found the incorporation of the autograft was slightly better than the allograft. However, the morbidity and pain suffered by the patients was far greater in the autograft group. Data from our institution as reviewed by Robert Molinari, M.D. and presented at the SRS, NASS and the Orthopaedic Academy showed that 66 out of 67 fresh frozen allografts showed remodeling and incorporation at a minimum of 5 years postoperatively without any incidence of graft collapse. These cases were all supplemented with posterior instrumentation and autogenous grafting. Another paper from our institution investigated by Kevin Eck, M.D. reviewed results of Harms titanium cages as anterior support. There was an extremely low incidence of cage extrusion, substantial cage subsidence, and cage collapse. However, judging an anterior fusion around the cages is very difficult both by plain radiographs and by CT evaluation.
Does a long fusion to the sacrum relieve back pain? Does a long fusion to the sacrum improve function? We are presently analyzing these questions at our institution. Preliminary answers are that long fusion to the sacrum does relieve back pain in a majority of the patients (85 %) if a solid fusion is achieved with excellent coronal and sagittal balance. Although mobility is lost, patients will report on increase in function if they are converted from a positive sagittal balance to neutral or negative sagittal balance.