Abstract
Background: A primary concern after posterior lumbar spine
arthrodesis is the potential for adjacent segment degeneration cephalad or
caudad to the fusion segment. There is controversy regarding the subsequent
degeneration of adjacent segments, and we are aware of no long-term studies
that have analyzed both cephalad and caudad degeneration following posterior
arthrodesis. A retrospective investigation was performed to determine the
rates of degeneration and survival of the motion segments adjacent to the site
of a posterior lumbar fusion.
Methods: Two hundred and fifteen patients who had undergone
posterior lumbar arthrodesis were included in this study. The study group
included 126 female patients and eighty-nine male patients. The average
duration of follow-up was 6.7 years. Radiographs were analyzed with regard to
arthritic degeneration at the adjacent levels both preoperatively and at the
time of the last follow-up visit. Disc spaces were graded on a 4-point
arthritic degeneration scale. Correlation analysis was used to determine the
contribution of independent variables to the rate of degeneration.
Survivorship analysis was performed to describe the degeneration of the
adjacent motion segments.
Results: Fifty-nine (27.4%) of the 215 patients had evidence of
degeneration at the adjacent levels and elected to have an additional
decompression (fifteen patients) or arthrodesis (forty-four patients).
Kaplan-Meier analysis predicted a disease-free survival rate of 83.5% (95%
confidence interval, 77.5% to 89.5%) at five years and of 63.9% (95%
confidence interval, 54.0% to 73.8%) at ten years after the index operation.
Although there was a trend toward progression of the arthritic grade at the
adjacent disc levels, there was no significant correlation, with the numbers
available, between the preoperative arthritic grade and the need for
additional surgery.
Conclusions: The rate of symptomatic degeneration at an adjacent
segment warranting either decompression or arthrodesis was predicted to be
16.5% at five years and 36.1% at ten years. There appeared to be no
correlation with the length of fusion or the preoperative arthritic
degeneration of the adjacent segment.
Level of Evidence: Prognostic study, Level IV (case
series). See Instructions to Authors for a complete description of levels of
evidence.
The prevalence of lumbar arthrodesis has continued to increase because of
the emergence of newer techniques of spinal instrumentation and improved
imaging modalities that allow for accurate recognition of spinal
abnormalities. The levels involved in the arthrodesis typically are
degenerative or unstable, and the ultimate goals are to provide relief of
symptoms and to restore stability. Retrospective studies on scoliosis as well
as longitudinal studies on lumbar fusion have suggested that lower lumbar
fusions predispose patients to problems in the adjacent motion
segments1-14.
Additionally, evidence of increased motion of cephalad adjacent segments and
increased disc compression at adjacent motion segments has been well described
in cadaveric
studies15-17.
Although adjacent segment degeneration in the lumbar and lumbosacral spine
has been examined extensively in previous biomechanical and clinical studies,
we are aware of no study that has specifically addressed the rate of
degeneration of adjacent segments. In addition, previous studies have not
demonstrated an association between radiographic evidence of degeneration of
adjacent segments and the long-term clinical outcome of posterior lumbar
fusion. Radiographic signs of degeneration of disc spaces adjacent to the site
of a lumbar fusion may reflect the natural history of lumbar spondylosis and
may only be meaningful when they are associated with clinical symptoms of
radiculopathy, discogenic pain, or stenosis referable to that
level18.
The objectives of the present study were to estimate the incidence,
prevalence, and rate of degeneration of the adjacent segments in the lumbar
spine following posterior lumbar arthrodesis, both radiographically and
symptomatically, and to determine which lumbar segments are at the greatest
risk for new symptoms. We also assessed whether multiple-level fusion is a
risk factor for adjacent segment disease and analyzed the correlation between
radiographic degeneration and findings warranting additional operative
intervention. Finally, we examined and analyzed the independent demographic
and surgical factors that were associated with clinical outcomes.
Between April 1983 and August 1994, 215 patients who had had a posterior
lumbar arthrodesis were evaluated. The hospital records, office charts, and
radiographs were reviewed and analyzed by an independent observer (G.G.) to
determine demographic characteristics, symptoms, preoperative and
postoperative diagnoses, and patient function at each follow-up visit. One
hundred and sixty-five of the 215 patients had had a posterior lumbar
intertransverse process arthrodesis that had been performed by the senior
author (E.G.D.) at a single institution for the treatment of degenerative
disease of the lumbar spine during this time-period. None of these patients
had an acute fracture or dislocation, had been managed for a neoplasm, or were
scheduled to have an additional anterior surgical procedure.
The remaining fifty patients had had either a previous posterior lumbar
arthrodesis at an outside institution or a remote arthrodesis that had been
performed by the senior author before 1983. These patients were included in
the survivorship analysis and were valuable for providing long-term data
points for the evaluation of disease-free survivorship based on the date of
the index procedure. Such patients were only included in the study if they had
had radiographic evidence of a healed lumbar fusion after the index
arthrodesis. Patients were not included in the radiographic analysis if
preoperative radiographs from the time of the index arthrodesis were not
available. This subgroup included twenty-nine female patients and twenty-one
male patients. The average age of these fifty patients at the time of the
index procedure was 41.9 years, and the average duration of follow-up was 13.2
years. Thirty of these fifty patients had a subsequent procedure at segments
adjacent to the site of the index fusion; specifically, twenty-seven patients
had an arthrodesis and three had a decompression.
The indications for the index posterior lumbar arthrodesis included (1)
progressive spondylolisthesis, (2) degenerative or iatrogenic
spondylolisthesis in patients undergoing decompression for the treatment of
spinal stenosis, (3) progressive lumbar scoliosis, (4) iatrogenic instability
resulting from an extensive decompression, (5) the occurrence of two or more
episodes of disc herniation at the same level, and (6) incapacitating
nonradicular back pain after the failure of nonoperative treatment. The
arthrodesis was performed at all levels associated with clinical signs and
symptoms and at which neural element compression was demonstrated on
neuroradiographic images. The diagnosis of adjacent segment disease was based
on the presence of instability, radiculopathy, or spinal stenosis that was
symptomatic enough for the patient to elect revision surgery. The criteria for
arthrodesis as opposed to decompression at an adjacent segment were the same
as those previously listed for the index arthrodesis.
All patients returned for regular postoperative visits that involved a
radiographic assessment and an examination by the senior author. The
persistence of symptoms, work status, functional status, the use of pain
medication, and the findings of a complete neurological examination were
documented. The outcome at each follow-up visit was rated as excellent, good,
fair, or poor on the basis of a modified function scale
(Table
I)14,19.
The category that was assigned was determined on the basis of the worst
outcome parameter.
The study group included 126 female patients and eighty-nine male patients.
The average age of the patients at the time of the index procedure was fifty
years (range, thirteen to eighty-five years). All 215 patients had a clinical
visit with documentation of function at least one year after the index
procedure. The average duration of follow-up was 6.7 years (range, one to
forty-one years). One hundred and seventy-eight patients had at least two
years of follow-up (average, 7.7 years). One hundred and ten arthrodeses (51%)
were performed with instrumentation, and 105 (49%) were performed without
instrumentation. Ninety-eight patients had a single-level arthrodesis. One
hundred and seventeen patients had a multiple-level arthrodesis; specifically,
seven patients had a five-level arthrodesis, eighteen had a four-level
arthrodesis, eighteen had a three-level arthrodesis, and seventy-four had a
two-level arthrodesis. Six of the seven five-level arthrodeses were from the
thoracic spine to the L5 vertebra.
Standard biplanar anteroposterior, lateral, flexion, and extension
radiographs of the lumbosacral spine from the preoperative visit as well as
from the last postoperative visit were reviewed for each patient. Lateral
radiographs demonstrating neutral, flexion, and extension views were measured
for anteroposterior translation and intervertebral disc height at each lumbar
segment. Objective intervertebral disc heights were measured with use of
published
methods20. The
degenerative grade at each lumbar disc level was rated at the time of the
index procedure and again at the time of the last radiographic follow-up
visit. These measurements were performed independently by two of the authors
(G.G. and J.C.W.). The amount of lumbar degeneration was classified, according
to the University of California at Los Angeles grading scale, as no disease
(Grade I), mild disease (Grade II), moderate disease (Grade III), or severe
disease (Grade IV) (Table
II)5.
Radiographic evidence of instability was defined, on the basis of published
standards, as >4 mm of translation or >10° of angular motion between
adjacent end plates on lateral flexion and extension radiographs when compared
with the adjacent cephalad and caudad
levels21.
Statistical Analysis
The incidence and prevalence of surgical intervention for adjacent segment
disease were calculated for each year, and a Kaplan-Meier survivorship curve
with 95% confidence intervals was constructed. Incidence was defined as the
percentage of patients who had not had revision surgery at the start of a
given year and had had subsequent development of new disease that was treated
surgically during that year. Prevalence was defined as the overall percentage
of patients who had surgery at an adjacent segment during a given
time-period.
Where applicable, two cephalad and two caudad lumbar motion segments
adjacent to the fusion were considered at risk for new disease. The prevalence
of symptomatic adjacent segment disease was calculated by dividing the number
of cases of new disease at that segment by the total number of segments at
risk for disease.
A Cox proportional-hazards model was used to determine the independent
variables that contributed to the rate of adjacent segment degeneration.
Independent variables included age at the time of the index procedure, gender,
preoperative diagnosis, the length of the fusion instrumentation, and the
length of time between the index procedure and the last follow-up visit. The
Fisher exact test was used to compare the preexisting disc degeneration at
adjacent levels with the development of adjacent segment disease. The level of
significance was set at p < 0.05.
Fifty-nine (27.4%) of the 215 patients had adjacent segment disease that
was symptomatic enough for them to elect to have a surgical procedure at the
adjacent level. Forty-four of these fifty-nine patients had a decompression
and arthrodesis, and fifteen had a decompression only. Postoperatively, new
disease at an adjacent level developed at a relatively constant rate of 3.9%
per year (95% confidence interval, 2.8% to 5.1%; range, 0% to 6.1%)
(Fig. 1).
Kaplan-Meier survivorship analysis was performed in order to assess the
rate of disease-free survival for the entire series of patients and to take
into account patients who had been lost to follow-up
(Fig. 2). The estimated rate of
disease-free survival was 83.5% (95% confidence interval, 77.5% to 89.5%) at
five years after the index operation and 63.9% (95% confidence interval, 54.0%
to 73.8%) at ten years after the index operation. This finding suggests that
16.5% of all patients who have had a posterior lumbar fusion will have new
disease warranting a second procedure at an adjacent level within the first
five years after the index procedure and that 36.1% will have new disease
within the first ten years after the index procedure.
There were significant differences among the various motion segments with
regard to the relative risk of adjacent segment disease (p < 0.001)
(Table III). The relative risk
of new-onset disease at the L2-L3 interspace was 2.7 times lower than that at
the L4-L5 level, which had the highest prevalence. The L5-S1 interspace had a
relatively low risk of subsequent degeneration, with a prevalence of 7.2%.
The 122 patients who did not have additional surgical intervention at an
adjacent segment and who had at least two years of follow-up were evaluated
with use of established criteria for
outcome14,19.
The average duration of follow-up (and standard error) from the time of the
index procedure was 6.5 ± 4.8 years (range, 2.0 to 38.1 years).
Fourteen of these 122 patients had an excellent outcome, forty-two had a good
outcome, forty-nine had a fair outcome, and seventeen had a poor outcome at
the last follow-up visit.
On the basis of the data in Table
II, each grade was assigned a numeric point value. The average
radiographic score for adjacent segment disease this 4-point scale was 1.81
points preoperatively and 2.24 points postoperatively. There was significant
degeneration at all levels when the most recent radiographs were compared with
the preoperative radiographs (Table
IV).
Cox regression analysis showed no significant correlation between adjacent
segment disease and diagnosis (p = 0.34), age at the time of surgery (p =
0.13), gender (p = 0.92), or instrumentation (p = 0.47), with the numbers
available. The Fisher exact test showed a trend but no significant correlation
between preexisting radiographic degeneration and adjacent segment disease (p
= 0.115). Contrary to our hypothesis, segments that were adjacent to a
single-level fusion had a three times higher risk for the development of
disease than did those that were adjacent to a multiple-level fusion (Cox
proportional hazards model, 3.4; 95% confidence interval, 1.83 to 6.23; p <
0.001).
The continued degeneration of motion segments adjacent to lumbar spinal
fusions is a potential concern for both patients and surgeons and accounts for
a substantial percentage of revision spine surgery. Although the development
of adjacent segment degeneration can be considered part of the normal aging
and degenerative process, this phenomenon appears to be at least partly
influenced by the altered stresses that arise as a consequence of lumbar
fusion15,17,18,22.
There have been many clinical studies, ranging in size from forty-five to
312 patients, that have described accelerated degeneration of lumbar segments
adjacent to the site of a previous
arthrodesis2,7,9-14.
Those studies have detailed disc-space narrowing and spondylolisthesis in the
adjacent segments after lumbar or, more commonly, lumbosacral fusion. The
reported prevalence of degeneration at the adjacent segments has ranged from
5% to
43%3,7,11.
However, the prevalence of lumbar surgery performed for the treatment of this
degeneration has been much lower (range, 2% to 15%). The majority of
subsequent operations have involved neural decompression rather than cephalad
extension of the fusion. The disparity in these data can be explained by
variability in the duration of follow-up and inconsistency in the definition
of adjacent segment disease in the various studies.
In addition, biomechanical studies have supported the increased prevalence
of degenerative disease adjacent to the
fusion15-17.
The authors of those studies postulated that increased stress or hypermobility
at the adjacent segment was a possible etiology of adjacent segment
degeneration. Lee and Langrana showed that there is increased stress at the
adjacent facet joints of L3-L4 and L4-L5 after lumbosacral
arthrodesis16.
Quinnell and Stockdale specifically addressed the influence of a single lumbar
floating arthrodesis on the rest of the lumbar spine and concluded that the
disc cephalad to the fusion is unaffected in terms of its external dimensions
whereas the discs caudad to the fusion exhibit a change in their loading
characteristics17.
More recently, Axelsson et al., in an in vitro model, assessed adjacent
segments with use of roentgen stereophoto-grammetric analysis and found
relative hypermobility in the juxtafused
segment23. The
findings of those biomechanical studies suggest that lumbar fusions produce
adverse consequences on the adjacent motion segments.
In the present study, the rate of surgical intervention for adjacent
segment disease was 3.9% per year during the first ten years following primary
posterior lumbar arthrodesis. Kaplan-Meier survivorship analysis predicted
that, at ten years, 36.1% of the patients would have sufficient disease to
warrant additional surgical intervention. We believe that our model accurately
describes the risk of symptomatic adjacent segment degeneration because it
takes into account patients who had died and those who had been lost to
follow-up.
We hypothesized that, after a multiple-level fusion, more motion would be
transferred to the adjacent segments, thus leading to a more rapid onset of
disc degeneration and new disease at the adjacent levels. Contrary to that
hypothesis, patients who had a multiple-level fusion were significantly less
likely to have symptomatic adjacent segment disease than those who had a
single-level fusion (p < 0.001). This finding may be explained by the fact
that a patient who has a single-level fusion has more levels at risk than a
patient who has a long fusion segment. For example, a patient who has a fusion
at L3-L4 has four levels at risk for adjacent segment degeneration, whereas a
patient who has instrumentation from the thoracic spine to L5 has only one
level at risk for disease.
Our radiographic data showed a large amount of inconsistency when the
intervertebral disc height was measured quantitatively. This finding was
concordant with the findings of other investigators who have attempted to
measure intervertebral disc height objectively but have found that it is
impossible to do so unless one carefully controls the tube-target-film
relationship, uses optimum radiographic techniques that include osseous
landmarks, and compensates for radiographic
magnification20,23.
Therefore, we developed a modified arthritis-grading scale to grade disc
degeneration qualitatively (Table
II).
The radiographic findings of the present study did show a significant
progression of the arthritic grade of the adjacent segment. However, the
clinical importance of this radiographic progression is undetermined. It is
expected that arthritic degeneration of a motion segment will progress with
time, regardless of whether or not the motion segment is adjacent to a fused
segment. The radiographic findings of the present study suggest that there is
no predictable way to determine which segments will degenerate in the future.
Although symptomatic degeneration is multifactorial, it may be less likely to
occur in arthritic segments than in mobile segments without arthritis.
Our study had several important limitations. The first limitation is that
the study was a retrospective review of a heterogeneous patient population.
Although a multivariate regression analysis was performed to assess the
contributions of independent variables such as age, gender, instrumentation,
and preoperative diagnosis, a more homogenous population might have provided a
stronger correlation between independent variables and adjacent segment
disease.
Another limitation of our study is that it lacked a control group for
comparison. A matched population of patients with spondylosis who refused
operative intervention would provide the most ideal control group if they were
available for long-term follow-up. We would then be able to assess the
degenerative changes in the adjacent segments without the influence of a
juxtaposed fusion. Instead, we have assumed that the adjacent segment disease
is a direct result of the surgical fusion.
The greatest limitation of our study is that reoperation was used as the
end point for survivorship analysis. This definition did not consider patients
with adjacent segment degeneration who might have benefited from surgical
intervention but did not have an operation because of unknown reasons such as
comorbidities or other unresolved variables. Pain is a complex matter that
involves psychosocial factors, which are likely to be more important than
physical factors. Of the patients who do choose surgery, it is never known how
they would fare without it. With this in mind, the survivorship data presented
in the present study most likely overestimated the percentage of disease-free
survival.
The present study offers insight into the natural history of a lumbar
fusion and estimates the rate of adjacent segment degeneration at five and ten
years after posterior lumbar arthrodesis. Our clinical findings did not
support a relationship between preoperative arthritic degeneration, the use of
instrumentation, or the length of fusion and subsequent degeneration of
adjacent motion segments. Although one cannot discern the contribution of
fusion apart from natural history, our data provide important information
regarding adjacent segment disease. The present study should provide useful
information for both the patient and the clinician and should guide future
research regarding adjacent segment disease.
Cochran T, Irstam L, Nachemson A.
Long-term anatomic and functional changes in patients with adolescent
idiopathic scoliosis treated by Harrington rod fusion.
Spine.1983;8:
576-84.8576
1983
[PubMed][CrossRef]
Brodsky AE. Post-laminectomy and
post-fusion stenosis of the lumbar spine. Clin Orthop.1976;115:
130-9.115130
1976
[PubMed]
Frymoyer JW, Hanley E, Howe J,
Kuhlmann D, Matteri R. Disc excision and spine fusion in the management of
lumbar disc disease. A minimum ten-year followup.
Spine.1978;3:
1-6.31
1978
[PubMed][CrossRef]
Hambly MF, Wiltse LL, Raghavan N,
Schneiderman G, Koenig C. The transition zone above a lumbosacral fusion.
Spine.1998;23:
1785-92.231785
1998
[PubMed][CrossRef]
Ghiselli G, Wang JC, Hsu WK, Dawson
EG. L5-S1 segment survivorship and clinical outcome analysis after L4-L5
isolated fusion. Spine.2003;28:
1275-80.281275
2003
[PubMed]
Harris RI, Wiley JJ. Acquired
spondylolysis as a sequel to spine fusion. J Bone Joint Surg
Am.1963;45:
1159-70.451159
1963
Lehmann TR, Spratt KF, Tozzi JE,
Weinstein JN, Reinarz SJ, el-Khoury GY, Colby H. Long-term follow-up of
lower lumbar fusion patients. Spine.1987;12:
97-104.1297
1987
[PubMed][CrossRef]
Lee CK. Accelerated degeneration
of the segment adjacent to a lumbar fusion. Spine.1988;13:
375-7.13375
1988
[PubMed][CrossRef]
Leong JC, Chun SY, Grange WJ, Fang
D. Long-term results of lumbar intervertebral disc prolapse.
Spine.1983;8:
793-9.8793
1983
[PubMed][CrossRef]
Miyakoshi N, Abe E, Shimada Y,
Okuyama K, Suzuki T, Sato K. Outcome of one-level posterior lumbar
interbody fusion for spondylolisthesis and postoperative intervertebral disc
degeneration adjacent to the fusion. Spine.2000;25:
1837-42.251837
2000
[PubMed][CrossRef]
Penta M, Sandhu A, Fraser RD.
Magnetic resonance imaging assessment of disc degeneration 10 years after
anterior lumbar interbody fusion. Spine.1995;20:
743-7.20743
1995
[PubMed][CrossRef]
Pihlajamaki H, Bostman O, Ruuskanen
M, Myllynen P, Kinnunen J, Karaharju E. Posterolateral lumbosacral fusion
with transpedicular fixation: 63 consecutive cases followed for 4 (2-6) years.
Acta Orthop Scand.1996;67:
63-8.6763
1996
[PubMed][CrossRef]
Schlegel JD, Smith JA, Schleusener
RL. Lumbar motion segment pathology adjacent to thoracolumbar, lumbar, and
lumbosacral fusions. Spine.1996;21:
970-81.21970
1996
[PubMed][CrossRef]
Whitecloud TS 3rd, Davis JM, Olive
PM. Operative treatment of the degenerated segment adjacent to a lumbar
fusion. Spine.1994;19:
531-6.19531
1994
[PubMed][CrossRef]
Yang SW, Langrana NA, Lee CK.
Biomechanics of lumbosacral spinal fusion in combined compression-torsion
loads. Spine.1986;11:
937-41.11937
1986
[PubMed][CrossRef]
Lee CK, Langrana NA. Lumbosacral
spinal fusion. A biomechanical study. Spine.1984;9:
574-81.9574
1984
[PubMed][CrossRef]
Quinnell RC, Stockdale HR. Some
experimental observations of the influence of a single lumbar floating fusion
on the remaining lumbar spine. Spine.1981;6:
263-7.6263
1981
[PubMed][CrossRef]
Boden SD, Davis DO, Dina TS, Patronas
NJ, Wiesel SW. Abnormal magnetic-resonance scans of the lumbar spine in
asymptomatic subjects. A prospective investigation. J Bone Joint
Surg Am.1990;72:
403-8.72403
1990
Hilibrand AS, Carlson GD, Palumbo MA,
Jones PK, Bohlman HH. Radiculopathy and myelopathy at segments adjacent to
the site of a previous anterior cervical arthrodesis. J Bone Joint
Surg Am.1999;81:
519-28.81519
1999
Pope MH, Wilder DG, Matteri RE,
Frymoyer JW. Experimental measurements of vertebral motion under load.
Orthop Clin North Am.1977;8:
155-67.8155
1977
[PubMed]
Wiltse LL, Winter RB. Terminology
and measurement of spondylolisthesis. JBone Joint Surg
Am.1983;65:
768-72.65768
1983
Torgerson WR, Dotter WE.
Comparative roentgenographic study of the asymptomatic and symptomatic lumbar
spine. J Bone Joint Surg Am.1976;58:
850-3.58850
1976
[PubMed]
Axelsson P, Johnsson R, Stromqvist
B. The spondylolytic vertebra and its adjacent segment. Mobility measured
before and after posterolateral fusion. Spine.1997;22:
414-7.22414
1997
[PubMed][CrossRef]