Imaging

Plain radiographic evaluation may demonstrate loss of disc height, disc osteophytes, facet arthrosis, spondylolisthesis, or scoliosis ( Figs. 36-1 A and 36-2 A ). In addition, the type and location of prior instrumentation may be identified. Computed tomographic (CT) images of the spine will confirm presence of a solid fusion or may identify a pseudarthrosis. Magnetic resonance imaging (MRI) can identify areas of central canal, lateral recess, or foraminal stenosis, as well as the degenerative status of the intervertebral discs (see Fig. 36-2 B ). Artifact from instrumentation may limit visualization on MRI. In this setting, postmyelogram CT imaging (see Fig. 36-1 B ) can improve visualization of the spinal elements and spinal canal.

A 56-year-old woman has neurogenic claudication 4 years after L3-5 decompression and fusion. A, Lateral radiograph demonstrates prior instrumentation at L3-5, as well as loss of disc height and retrolisthesis at the L2-3 level. B, Postmyelogram computed tomographic images demonstrates dye cutoff and stenosis at the L2-3 level.

A 62-year-old woman. A, Lateral radiograph demonstrates loss of disc space height at both the L3-4 and L5-S1 levels with spondylolisthesis at L5-S1. B, Sagittal T2-weighted magnetic resonance imaging identifies stenosis at the L3-4 level. C, Postoperative lateral radiograph after combined anterior-posterior fusion at L3-S1 with posterior decompression at the L3-4 level.

Incidence

Adjacent segment degeneration is classified according to presence on radiographic evaluation or disease, or both, that results in clinical findings. In Park et al.’s review, the radiographic incidence rate of adjacent segment degeneration varied substantially from 5.2% to 100% during average postoperative observation ranging from 36 to 369 months. It is important to recognize that radiographically apparent, asymptomatic, adjacent segment degeneration is common but does not necessarily correlate with functional outcomes.

The incidence rate of symptomatic adjacent segment disease was lower, ranging from 5.2% to 18.5% during 44.8 to 164 months of observation. In two series, the mean time to symptomatic adjacent segment degeneration requiring surgery ranged from 7 to 13 years. Ghiselli et al. performed survivorship analysis to describe the degeneration of the adjacent motion segments on 215 patients after posterior lumbar arthrodesis with a mean follow-up of 7 years. Kaplan–Meier analysis predicted a disease-free survival rate of 83.5% (95% confidence interval, 77.5–89.5%) at 5 years and 63.9% (95% confidence interval, 54.0–73.8%) at 10 years after the index operation. The rate of symptomatic degeneration at an adjacent segment warranting either decompression or arthrodesis was predicted to be 16.5% at 5 years and 36.1% at 10 years. This would approximate a rate of adjacent segment disease at 3% per year.

Age and Other Patient Characteristics

Although no studies directly compare the impact of age on the development of adjacent segment degeneration, several authors have reported adjacent segment disease in older patients. Park et al. note that progressive ALD with age is a major contributor. Rahm and Hall note that the incidence of adjacent segment degeneration was associated with increasing patient age. Chou et al. performed a retrospective review of 32 patients older than 60 years after lumbar fusion with a follow-up of at least 4 years. Three patients each with translation greater than 4 mm in adjacent segments above and below were found in both the short and long (≥3 segments) fusion groups, and the overall incidence rate of adjacent segment disease was 18.7%, greater than the reported rates among all ages. The authors did not state how many of these patients were symptomatic.

Etebar and Cahill evaluated 125 patients after instrumented lumbar fusion for the treatment of degenerative instability to determine risk factors for adjacent degeneration. The mean follow-up period for this group was 44.8 months. Of a total of 18 patients who experienced development of symptomatic next-segment degeneration at a previously asymptomatic level, 15 were postmenopausal women. Etebar and Cahill conclude that the risk for adjacent segment failure appeared to be greater in postmenopausal women. Also, they note that 20% of all patients with next-segment failure were cigarette smokers, suggesting another potential factor. Cheh et al. have retrospectively reviewed 188 patients with a minimum 5-year follow-up period after lumbar or thoracolumbar fusion. The authors note that patients older than 50 years were at statistically significant greater risk for development of clinical disease.

Preexisting Degeneration

Ghiselli et al. evaluated 215 patients who had undergone posterior lumbar arthrodesis at an average duration of follow-up of 6.7 years. The authors note a trend toward progression of preoperative arthritic grade at the adjacent disc levels. No significant correlation existed, however, between the preoperative arthritic grade and the need for additional surgery. In addition, Okuda et al. have evaluated 87 patients after underwent posterior instrumented lumbar interbody fusion (PLIF) at L4-5 and note that preoperative radiographic findings did not correlate with postoperative degeneration.

Hambly et al. reviewed 42 patients with an average follow-up period of 22.6 years who had previously undergone a posterolateral lumbar or lumbosacral fusion and compared these patients with control subjects who had no surgery. Interestingly, degenerative changes occurred at the second level above the fused levels with a frequency equal to those occurring in the first level. No statistically significant differences were found between the study group and the cohort group with regard to radiographic changes within the transition zone (the cephalad two motion segments). Radiographic changes occurred within the transition zone cephalad to a lumbar or lumbosacral fusion; however, these changes are also seen at the same levels in control subjects who had no surgery.

Throckmorton et al. performed a retrospective review of 25 patients treated with lumbar fusions for degenerative instability to assess the effect of preoperative adjacent segment degeneration. Interestingly, they demonstrate no adverse impact on clinical outcomes when the lumbar fusion ended adjacent to a degenerative motion segment in comparison with a normal intervertebral disc, but they note that larger prospective studies are needed evaluate whether there is a benefit to including degenerated adjacent segments in a lumbar fusion for degenerative instability.

Willems et al. evaluated 82 patients who underwent a lumbar fusion with preoperative discography adjacent to the level of fusion. They found no difference in outcomes between those patients with symptomatic discs adjacent to the fusion and those with normal adjacent discs as determined by discography, suggesting that preoperative disc status, as determined by provocative discography, may not predict postoperative outcomes.

Iatrogenic Injury

Operative factors during the index surgery may result in increased risk for junctional level degeneration. Lai et al. have evaluated the integrity of the posterior spinous complex between the fused segment and motion segment, and its impact on the incidence of ALD and instability in 101 patients treated with instrumented, posterolateral lumbar fusion and followed up for at least 6 years. They note that damaging the integrity of the posterior complex between the fused segments and the neighboring motion segments may jeopardize lumbar spine stability, and that sacrificing either the supraspinous ligament or the tendon insertion points on the spinous processes might have lead to an accelerated development of adjacent instability. Interestingly, other authors advocate preventive decompression of the adjacent segment, or selective decompression without fusion given the risk for later occurrence of adjacent segment stenosis after multisegment posterolateral fusion (PLF).

The integrity of the adjacent segment facet joints must be preserved to prevent the risk for the development of future instability. One study radiographically evaluated the incidence of adjacent superior segment facet joint violation after transpedicular instrumentation by noting the location of pedicle screws in relation to the facet joints in 106 patients using CT and plain radiographs. Facet joint violation occurred in more than 30% of the patients and 20% of the screws in this study. This violation suggests a cause of long-term deterioration in the clinical results and development of ALD after transpedicular instrumentation. This is further supported by studies that demonstrate that the rate of symptomatic adjacent segment disease is greater in patients with transpedicular instrumentation (12.2–18.5%) compared with patients fused with other forms of instrumentation or with no instrumentation (5.2–5.6%).

Number of Fusion Levels

In several studies, multilevel fusions were more likely to result in clinically significant ALD. Shone et al. found that as segmental spinal instrumentation progressed from one level to three levels, the overall torsional and flexural rigidity of the system increased, and that changes in the motion pattern of the upper intact motion segments became more distinct as the fixation range became more extensive and as the rigidity of the construct increased. In a biomechanical study, Sudo et al. found that as the number of fused segments increased, the intradiscal pressure and lamina strain in the upper adjacent segment also increased.

Greiner-Perth et al. evaluated 1680 patients who underwent a PLIF with a mean follow-up period of 5 years. The most important, statistically significant difference between the multisegmental PLIFs and the monosegmental or bisegmental PLIFs was the rate of adjacent segment decompensation (5.1% vs. 2.3%). Although more multisegmental PLIFs required reoperation compared with monosegmental or bisegmental PLIFS, a significant difference was not found in the reoperation rate. Aiki et al. have evaluated different factors that were associated with need for reoperation for ALD in 117 patients who had undergone posterior lumbar fusion and were followed for a minimum of 7 years. Of factors including sex, age, initial pathologic condition, and initial spinal fusion and decompression methods, only multilevel fusion was associated with a greater rate of reoperation, and it reached statistical significance ( P < 0.04). Cheh et al. have identified length of fusion as a significant risk factor in the development of radiographic adjacent segment degeneration in their retrospective review of 188 patients.

Other studies, in contrast, have demonstrated no differences in the development of ALD after multilevel fusion. Chou et al. did not find a difference in the development of ALD for short versus long fusion when radiologically and clinically evaluating 32 patients who underwent PLF with 4 years of follow-up. The incidence rate was 16.7% (3/18) in the short fusion group and 21.4% (3/14) in the long fusion group. Ghiselli et al. also found no correlation with the length of fusion and development of degeneration of the adjacent segment. In a cadaveric, biomechanical comparison of adjacent segment motion at L3-4 for an L4-5 versus an L4-S1 fusion model, extending a fusion across L5 to the sacrum did not increase motion at the L3-4 level. The authors suggest that extending a fusion to the sacrum in a lower lumbar fusion should not result in increased adjacent segment degeneration.

Instrumentation and Approach

The rigidity of instrumentation constructs provides an ideal environment for fusion but may be detrimental to the adjacent levels. Etebar and Cahill note that the rate of adjacent segment failure is clearly greater for patients in whom lumbar fusion with rigid instrumentation is performed to treat degenerative instability. Biomechanical studies demonstrate that instrumented constructs produced greater segmental displacement values at the upper residual intact motion segment when compared with those of the intact spine. In contrast, the instrumented constructs decreased their segmental displacement values at the lower residual intact motion segment with higher magnitude of the translational (shear) motion taking place compared with the intact spine in flexion/extension and lateral bending.

The use of unilateral fixation has been thought to minimize ALD by diminishing the possibility of iatrogenic injury and decreasing the construct rigidity. Fernandez-Fairen et al. demonstrate equivalent outcomes between unilateral instrumentation and bilateral instrumentation for lumbar spondylolisthesis when performed in addition to one- or two-level PLF. The authors found no significant differences for adjacent degeneration between unilateral and bilateral instrumentation.

The use of interbody fusion devices has been suggested to be another potential cause of adjacent segment degeneration. Sudo et al. have found increased intradiscal pressure and lamina strain in the upper adjacent segment with specimens subjected to the PLIF procedure compared with a PLF procedure. This increase was related to the fact that the range of motion at the reconstructed levels was significantly decreased in the PLIF models compared with that in the PLF alone models after both the one- and two-segment fusions.

Min et al. evaluated 48 patients with preoperative spondylolisthesis and minimal ALD who underwent instrumented L4-5 fusion and were divided into anterior lumbar interbody (ALIF) versus transforaminal lumbar interbody (TLIF) fusions. Radiographic progression of adjacent segment degeneration was found in 44.0% of the patients in the ALIF group and in 82.6% of those in the TLIF group ( P = 0.008), whereas clinical success rates were 92.0% and 87.0% in the ALIF and PLIF groups, respectively. This would suggest that posterior interbody fusion procedures may place patients at a greater risk for adjacent segment degeneration.

Okuda et al., in a retrospective study of 87 patients after PLIF at L4-L5 for degenerative spondylolisthesis, evaluated whether PLIF fusion accelerated degenerative changes at unfused adjacent levels. The authors demonstrate no correlation between radiologic degeneration of cranial adjacent segment and clinical outcomes. The rate of reoperation (4%) at 2 years matches the reported rates in PLF, suggesting the posterior interbody fusion may not have a detrimental effect on the adjacent levels.

Alignment

An in vitro biomechanical study of adjacent segment motion (at L3-4 and L5-S1) after a simulated lumbar interbody fusion of L4-5 in different sagittal alignments was performed. Hypolordotic alignment at L4-5 caused the greatest amount of flexion/extension motion at L3-4, and the differences were statistically significant in comparison with intact specimen, in situ fixation, and hyperlordotic fixation. Hyperlordotic alignment at L4-5 caused the greatest amount of flexion/extension motion at L5-S1, and the difference was statistically significant in comparison with intact specimen but not in situ fixation or hypolordotic fixation. Chen et al. performed a biomechanical study in which porcine lumbar spines were instrumented with pedicle screw implants from L2 to L4, and each specimen was tested in either lordotic, straight, or kyphotic alignment. The authors conclude that an iatrogenically produced kyphotic lumbar spine may cause greater motion at the adjacent, cranial segment as compared with a lordotic lumbar spine.

Oda et al. have evaluated the effects of spinal fusion and kyphotic deformity on the adjacent motion segments in a sheep model. The group fused in kyphosis showed increased compensatory hyperlordosis at cephalad and caudal to the fusion. Furthermore, the kyphotic PLF significantly influenced cephalad adjacent motion segment biomechanics by inducing more stiffness in the posterior ligamentous complex and increasing lamina strain under flexion/extension loading. Similarly, histologic evaluation demonstrated degenerative changes of the cephalad facet joints. Furthermore, compensatory hyperlordosis at the cephalad adjacent level leads to lordotic contracture of the posterior ligamentous complex. These results indicate that a kyphotic deformity may lead to facet joint contracture and facet arthritis, and may serve as the origin of low back pain at the cephalad adjacent level.

Other authors have demonstrated no effect of spinal alignment on adjacent level degeneration. Lai et al. have investigated the effect of postoperative lumbar sagittal alignment on the development of adjacent segment instability. The patients were retrospectively divided into two groups (hypolordotic and hyperlordotic) evaluating lumbar lordosis on lateral lumbosacral views. The authors conclude that restoration of lumbar lordosis during short segment fusion did not prevent the development of adjacent instability. Schlegel et al. reviewed 58 patients who underwent a thoracolumbar, lumbar, or lumbosacral fusion with an average symptom-free period of 13.1 years before presentation with severe symptoms, necessitating further surgery at the adjacent segment. The authors note that sagittal and coronal imbalances appeared to play a role in breakdown, although statistical significance was not evident. Pellise et al. found no relation to the risk for adjacent segment disease.

Caudal Degeneration

Although adjacent segment degeneration and resulting kyphosis occur more often at the cranial end of instrumented fusions, segmental degeneration and deformity may also occur at the caudal end. Kwon et al. have described a series of patients with progressive lumbar kyphosis or sagittal decompensation associated with caudal segment degeneration after an instrumented lumbar fusion. In 13 patients, the most common mode of caudal junctional decompensation was related to failure of the most distal fixation. Sagittal decompensation occurred even in the presence of satisfactory lumbar lordosis. The authors note that the potential for failure of L5 pedicle screws after an index surgery warrants consideration of extending fusions to the sacrum/ilium.

Edwards et al. have retrospectively reviewed 34 consecutive adult patients after fusion from the thoracic spine to L5 with an average follow-up of 5.6 years (2.1–14.3 years). The authors identify risk factors for caudal degeneration including preoperative positive sagittal balance, younger age, and preoperative radiographic degeneration. Similarly, Edwards et al. have observed loss of L5 fixation in 18% of patients as another cause of caudal decompensation.