High Grade L5-S1 Spondylolisthesis

C.P. presented as a 28–year–old male, diagnosed at age 13 as having spondylolisthesis. He was treated with a removable rigid lumbar orthosis (TLSO) briefly, at the time of his initial diagnosis and again at age 23. In each instance, as the low back pain (LBP) responded the patient discontinued the brace and did not follow up. Intermittently he would experience episodes of LBP.

In the two years preceding our initial evaluation, C.P. began to experience progressive paresthesia and ultimately pain in the left L5 and S1 distributions. Generally, the LBP was more severe and limiting than were the radicular symptoms. The back pain was strongly mechanical in nature: aggravated by prolonged sitting and standing; severely provoked by bending and twisting; and improved only by recumbency. The radicular symptoms varied with the severity of the lumbosacral pain. In the month prior to our first visit, perianal hypesthesia, penile numbness, and left gastrocnemius weakness developed.

There was no bowel or bladder incontinence.

Physical examination demonstrated an absence of the ankle reflexes bilaterally, marked numbness to light touch and pinprick in the left L5 and S1 dermatomes and the left perianal region, and 3/5 left gastrocnemius strength. Straight leg raising was positive at 30° on the left. Anal sphincter tone and voluntary control were intact.

Neurologically, the right lower extremity was intact with the exception of the absent ankle reflex. The spinal range of motion was provocative of LBP, particularly in extension, provoking both LBP and left L5–S1 radicular pain. The spinal rhythm was also abnormal.

Plain radiographs (Figs 1,2) demonstrate a high grade (grade 5) spondylolisthesis. MRI and myelographic evaluation (Figs 3,4) reveal the expected high–grade canal compromise at L5–S1. Somatosensory evoked potentials document a dense left S1 radiculopathy, though they were intact in the L5 dermatome.

Fig. 1	Fig. 2	Fig. 3	Fig. 4
Click on thumbnail for full view.

L5–S1 Spondylolisthesis: Discussion
Francis Denis, MD
University of Minnesota, Minneapolis, MN

The best management of this patient would have been an earlier fusion, sometime between age thirteen and the completion of his growth, so as to avoid altogether the neurological complications that he is presenting with. At this time, the involvement of the left L5, S1, S2 and (most likely) S3 roots suggests a severe central stenosis of the L5–S1 level as well as severe foraminal stenosis. The MRI cuts presented do not demonstrate severe left L5–S1 foraminal stenosis, but the plain films suggest it. The implication for involvement of the S1 and S2 roots resides in the significant weakness of the left gastrocnemius which does not occur with compression of the S1 root alone. In addition, the numbness extending all the way to the left perianal region suggests involvement of at least S3 and probably S4 and S5 on the left. The axial view on the MRI may have been relatively useless because the plane of maximal stenosis of the central canal is actually perpendicular to the plane of the axial cuts. For this reason it would be very useful to obtain some coronals to demonstrate the severe encroach–ment on the lumbosacral spinal canal, as it is most likely to exist on that plane.

In terms of management, surgery is indicated. The most important part of the operation will be an adequate decompression of the central spinal canal at L5–S1. It appears from some of the x–rays that, in addition to the L5 lamina, some of the lamina of L4 may have to be removed in order to obtain complete decompression. Secondly, decompression of the L5 nerve root in the left L5–S1 foramen is to be carried out. Insofar as the right L5 nerve root is concerned, I would also do a prophylactic foraminotomy to avoid the foraminal stenosis that frequently results from instrumentation of a spondylolisthesis.

I would recommend using an instrumented fusion with pedicle fixation at L4 and S1. I feel it would be very dangerous to try to insert a screw into the L5 pedicles, due to their size. The posterolateral fusion would include the transverse process of L5, which is very deep below the plane joining the transverse process of L4 and the sacra ala. When looking at the superior endplate of L4 in relationship to the back of S1, it is clear that the "slip angle" of L4 is essentially neutral, with perhaps even some lordosis.

Instrumentation without the addition of foraminal decompression may result in neurologic deficit, primarily in terms of the left L5 nerve root. This occurs typically when compression between screws is applied in the hope of obtaining some lordosis with instrumentation. Such a maneuver tends to close the foramen and further compress the L5 nerve root. Similarly, in the event of internal fixation of the spine in compression without central decompression, I would expect a cauda equina syndrome to be created postoperatively.

Although anterior fusions are frequently helpful in conjunction with the posterior stage, they tend to destabilize the lumbosacral level. The gain in terms of improved fusion is not significant in high–grade spondylolisthesis.

Discussion
Norbert Passuti, MD
C.H.R.U. Hôtel Dieu, Nantes, France

This case represents the most difficult situation we encounter of a dysplasic spondylolisthesis in an adult with the compression of L5 and S1 left roots and the beginning of neurologic deficit.

This is real spondyloptosis with severe lumbo–sacral kyphosis. We do not have dynamic x–rays in hyperextension but according to the MRI, we suppose that it is a fixed kyphosis with a high–grade canal compromise at L5–S1.

The different problems to solve are:
1. A large decompression at L5–S1, particularly a complete release of the L5 root very laterally around and below the L5 pedicle.
2. Correction of the lumbosacral kyphosis to obtain a segmental lordosis L4–L5–S1.
3. An anterior fusion for long–term stability.
4. Correction of the spondylolisthesis around grade 2 to obtain a solid construct for the postero–lateral fusion.

In this example we would begin with the anterior approach (transperitoneal or retroperitoneal). To control and protect the presacral nerve we would perform a large anterior release with a L5–S1 discectomy, resect the inferior part of L5 and the superior area of S1, and graft with cancellous bone.

During the same surgery we would also make a posterior approach. With hips and knees extended, we would perform a large posterior decompression (controlling the L5 roots and the L5 pedicles), and extend the fixation to L4 using screw instrumentation. A solid construct in S1 is needed with one convergent sacral screw in S1 and a divergent screw below (Stephanie connector) and an iliac extension.

In adults we can use a double–threaded screw at L5. Two short rods are fixed at S1 and slowly the rods are applied at L4 to correct the kyphosis. A moderate distraction between L4 and S1 is possible to allow the correction of the spondylolithesis by locking the two L5 screws.