Adjacent Level Disk Disease—Is it Really a Fusion Disease?

Adjacent segment degeneration (ASD) is a relatively common phenomenon after spinal fusion surgery. Whether ASD is a consequence of the previous fusion or an individual’s predisposition to continued degeneration remains unsolved to date. This article summarizes the existing biomechanical and clinical literature on the causes and clinical impact of ASD, as well as possible risk factors. Further, the theoretical advantage of motion-preserving technologies that aim to preserve the adjacent segment is discussed.

A century ago, Albee and Hibbs independently described spinal fusion to treat Pott’s disease and spinal deformity. Anecdotal case reports on adjacent segment degeneration (ASD) as an uncommon complication of lumbosacral fusion started to emerge a few decades later. Since the time of Albee and Hibbs, indications for spinal fusion have expanded considerably and, especially in the last 20 years, we have witnessed an increasing number of spine fusions performed for degenerative conditions of the lumbosacral spine. Operative treatment of these inherently benign conditions has evoked discussions about the long-term sequelae of spinal fusions. The fate of the adjacent segment after rigid fusion of the lumbosacral spine has been increasingly studied and reported.

Although the predisposing factors for developing adjacent segment problems after spinal fusion are largely unknown, altered biomechanics of the adjacent segments has been emphasized. Biomechanical studies have shown that spinal fusion increases intervertebral motion, intradiscal pressure, and facet joint stresses of the adjacent levels. The rationale behind the more recent motion-preserving technologies is to protect the adjacent levels from these adverse consequences, thus preventing the development of ASD by preserving motion of the operated levels. However, some recent biomechanical and clinical studies have suggested that the biomechanics of the adjacent segments after a fusion procedure may not be altered as much as previously thought and that, instead, ASD is a sign of continued progression of the degenerative process.

This article summarizes the existing literature on the ASD, both biomechanical and clinical, to provide the reader with our current understanding of the pathogenesis of and the possible risk factors for this increasingly common clinical problem. Moreover, the clinical impact of ASD is discussed. Finally, the question whether the current literature supports the theoretical rationale behind motion-preserving technologies that also aim to protect the adjacent segments will be addressed.

Incidence

The rate of radiologically verified ASD after lumbar and lumbosacral fusions reported in the literature since late 1980s ( Table 1 ) has varied from 11% to 100%. The true rate of ASD is difficult to define because of the retrospective nature of most of the studies and variable follow-up times. Moreover, the incidence of ASD varies according to the definition applied. Most clinical evidence suggests that the level cranial to the previous fusion is more susceptible to subsequent degeneration compared with levels below the fusion. This clinical observation has further been corroborated by in vitro biomechanical studies. Degenerative changes do not appear to be limited to the first adjacent cranial segment but have also been observed at multiple levels above a previous fusion.

Table 1

Clinical studies on ASD after a lumbosacral fusion

Author	Study Design	Number of Potential Patients	Number of Included Patients	Diagnosis	Type of Surgery	Age at Time of Surgery	Length of FU	Incidence of ASD	Incidence of Reoperations During FU	Significant Risk Factors
Lehmann et al, 1987	Retrospective	94	33	Mixed	Noninstrumented posterior fusion	NA	33 y	Stenosis 42% Instability 45%	15%	—
Kumar et al, 2001	Retrospective	54	28	DDD	Posterior fusion	NA	>30 y	Loss of disk height 36% Instability 14%	NA	—
Hambly et al, 1998	Retrospective	148	42	Mixed	Noninstrumented posterolateral fusion	NA	22.6 y	Disk space ossification 62% Facet joint arthrosis 52%	—	—
Remes et al, 2005	Retrospective	129	102	Isthmic spondylolisthesis	Noninstrumented posterior or posterolateral fusion	15.9 y	21.0 y	DDD 54% Loss of disk height 21% Facet joint degeneration 79%	—	—
Wai et al, 2006	Retrospective	64	39	Discogenic LBP	Noninstrumented ALIF	NA	20.5 y	74 % (advanced in 31%)	6%	—
Seitsalo et al, 1997	Retrospective	175	145	Isthmic spondylolisthesis	Posterior or posterolateral noninstrumented fusion	14.3 y	15.4 y	Loss of disk height 17–32 %	—	—
Disch et al, 2008	Retrospective	102	102	Isthmic spondylolisthesis, DDD	ALIF with or without posterior fusion	54 y	13.8 y	26%	12%	Floating L4/L5 fusion
Ekman et al, 2009	RCT	111	80	Isthmic spondylolisthesis	Posterolateral fusion with or without instrumentation	39 y	12.6 y	Disk height loss (UCLA grading scale) 38%	—	Concomitant laminectomy (UCLA criteria) Instrumentation NS
Schulte et al, 2007	NA	65	40	DDD, Isthmic spondylolisthesis	360° fusion	32.6 y (Isthmic) 45.1 y (DDD)	114 mo	Disk height reduction 11%–12% (Isthmic) 23%–25% (DDD)	—	Older age Multilevel fusion
Videbaek et al, 2010	RCT	148	95	DDD, Isthmic spondylolisthesis	ALIF + instrumented posterolatera fusion vs instrumented posterolateral fusion	45 y	8–13 y	89% with MRI evidence of ASD	9%	Older age
Cheh et al, 2007	Retrospective	188	188	Degenerative	Circumferential or posterior instrumented fusion	55 y	7.8 y	43% (56% symptomatic)	—	Age >50 y Multilevel fusion Level of proximal instrumented vertebra
Gillet, 2003	Retrospective	149	106	Degenerative	Instrumented posterolateral fusion	55 y	2–15 y	41%	20%	—
Ghiselli et al, 2003	Retrospective	NA	32	75% degenerative spondylolisthesis	Posterolateral fusion with or without instrumentation	56 y	7.3 y	NA	3%
Ahn et al, 2010	Retrospective	NA	3188	Mixed	45% PLIF 55% other	57 y	NA	80% (in the proximal segments)	3.5 %	Age >61 y Degenerative disease Multilevel fusion Male gender
Ghiselli et al, 2004	Retrospective	NA	215	Mixed	Posterolateral fusion with or without instrumentation	50 y	6.7 y	NA	28%	Single-level fusion Fusion level
Lai et al, 2004	Retrospective	107	101	Spondylolisthesis	Instrumented posterolateral fusion	61 y	6–7 y	23%	—	Decompression technique
Miyakoshi et al, 2000	Retrospective	74	45	Spondylolisthesis	PLIF + posterior fixation	58 y	6 y	100%	—	—
Kumar et al, 2001	Retrospective	NA	83	Degenerative disease	Mixed	51.6 y	5 y	36% above the fusion	17%	Abnormal sacral inclination
Chou et al, 2002	Retrospective	44	32	Degenerative	Decompression and posterolateral instrumented fusion	70.5 y	56 mo	Instability 19%	—	—
Min et al, 2008	Retrospective	NA	48	Mixed	Interbody fusion	51.6 y – 56.2 y	45 mo	63%	7%	Loss of preoperative lumbar lordosis Younger age
Etebar and Cahill 1999	Retrospective	125	125	Degenerative	Posterolateral instrumented fusion	NA	44.8 mo	Symptomatic ASD 14%	—	Multilevel fusion
Throckmorton et al, 2003	Retrospective	148	25	Degenerative	Posterior fusion	56	>2 y	Disk degeneration 80%	—	—
Okuda et al, 2004	Retrospective	NA	87	Degenerative spondylolisthesis	PLIF	64 y	43 mo	29%	4%	—
Aota et al, 1995	NA	72	65	Mixed	Mixed	55.8 y	39 mo	Instability 25%	—	Age >55 y
Kaito et al, 2010	Retrospective	97	85	Spondylolisthesis	PLIF + posterior fixation	64.1 y	38.8 mo	28%	13%	Distraction of the disk space
Park et al, 2007	Retrospective	132	34	Isthmic spondylolisthesis	PLIF + posterior fixation	48.9 y	24.7 mo	21%	0%	—

Abbreviations: ALIF, anterior lumbar interbody fusion; DDD, degenerative disk disease; FU, follow-up; LBP, low back pain; NA, not announced; NS, non-significant; PLIF, posterior lumbar interbody fusion; RCT, randomized controlled trial; UCLA, University of California, Los Angeles.

Symptomatic ASD (ie, adjacent segment disease) is relatively rare compared with radiologically verified degenerative changes of the adjacent levels after a lumbosacral fusion, with reported incidence between 0% and 28%. Ghiselli and colleagues performed a retrospective analysis of 215 fusion patients at an average 6.7 years after a posterior lumbar fusion with or without instrumentation. In their series, new symptoms related to ASD, severe enough to warrant reoperation, developed at a fairly constant rate of 3.9% per year starting from the first postoperative year. Disease-free survival (ie, no reoperation for ASD) was 83.5% and 63.9% at 5 and 10 years after the index surgery, respectively. Another analysis of 3,188 thoracolumbar fusion patients showed an annual decrease of 0.6% in the survival rate when reoperation was defined as the end point. Most clinical data suggest that functional outcomes after lumbar fusion surgery seem to be unaffected by radiographic, asymptomatic ASD, although some studies have reported less favorable outcomes in patients with ASD. A long-term follow-up of two randomized surgical groups correlated adjacent segment disk degeneration, foraminal compromise, and spinal stenosis to a worse clinical outcome.

Causes

To date, most investigators have emphasized the role of biomechanical alterations induced by lumbosacral fusion, namely increased intervertebral motion, intradiscal pressure, and facet joint loads at the adjacent segments, in the pathogenesis of ASD. The literature, however, provides controversial and conflicting evidence. The following will outline the present knowledge on the biomechanical behavior of the adjacent segments and correlate it to clinical data where available and applicable.

Adjacent Segment Intervertebral Motion

Changes induced by rigid fixation and simulated one-level to three-level fusion have been examined extensively using in vitro human cadaveric and animal models, as well as validated finite element models (FEM). Several of these studies have demonstrated significant changes in the intersegmental rotation of the first cranial segment above a simulated fusion, most notably in flexion-extension motion. Specifically, the percent increases of the motions of the first adjacent segment after a simulated fusion have ranged from 17% to 103% for flexion, from 17% to 67% for extension, from 6% to 94% for lateral bending, and from 6% to 20% for axial rotation. Some studies have shown that extending the length of the fixation leads to more distinct increases in the intervertebral motions. On the other hand, several studies have demonstrated only minor changes in the intervertebral rotations of the adjacent segment after a simulated fusion. With a specific hybrid testing protocol introduced by Panjabi, significant increases have been demonstrated, not only at the immediate adjacent level to a simulated fusion, but spread across multiple cranial levels.

The large variation observed in the intervertebral rotation of the adjacent segment is at least partly explained by different biomechanical testing modes. One mode is based on the assumption that, postoperatively, the fused patients will try to move their spines to the same extent as preoperatively. By definition, this would demand higher loads, which would then cause the remaining mobile segments to compensate for the lost motion of the fused segment by demonstrating increased range of motion. In biomechanics, this assumption would require displacement-controlled testing protocols and, specifically, a hybrid set-up for identification of adjacent segment changes. Another possibility is that, after a fusion procedure, the patient settles with the restricted lumbar motion, thus applying the same load to the unfused segments as before the operation. Biomechanically, this hypothesis could be tested by load-controlled protocols. In the first scenario, adjacent segment alterations after a fusion can be expected but, in the second one, they should be practically nonexistent.

From a biomechanical perspective, the best way to test for adjacent segment changes has not been worked out. The hybrid protocol is probably the best technique, but it has not been used extensively to date. Furthermore, it is of concern that some load-controlled experiments have observed significant displacement changes at adjacent levels, against the theoretical basis of this testing protocol (see previous discussion). Thus, proper interpretation of the results of biomechanical research calls for careful consideration of the methods used.

in vitro laboratory results are difficult to corroborate in in vivo conditions. In one such study on adult canines, Dekutoski and colleagues found increased motion at the adjacent segment to a spinal fusion with physiologic loading and concluded that the animals attempted to reproduce the preoperative range of lumbar motion. The results of this study would support the hypothesis that the adjacent levels after a spinal fusion are subjected to increased loads with physiologic loading conditions.

Lumbar spine in vivo kinematics in patients after spinal fusion operations have been examined using functional flexion-extension radiography, videofluoroscopy, and radiostereometry. Luk and colleagues compared the lumbar spine range of motion between clinically asymptomatic and completely pain-free patients with an anterior one-level or two-level fusion and asymptomatic volunteers using standardized flexion-extension radiographs. According to their results, the lumbar spine of the fusion patients was significantly less mobile than in the control subjects, which would suggest that the adjacent segments do not compensate for the lost motion at the fused segments. However, the percent contribution of the unfused segments to the total lumbar motion was increased because the total lumbar range of motion decreased proportionally more than that of the individual levels. Similar results regarding reduced total lumbar range of motion and no compensatory increase in the motion of the levels adjacent to a fusion have been reported by other investigators using flexion-extension radiography. Functional radiography on patients from the US Food and Drug Administration Investigational Device Exemption randomized controlled trials (RCTs) comparing fusion and disk arthroplasty show that fusion patients have slightly reduced total lumbar motion postoperatively with a significant increase in the segmental contribution of the first cranial adjacent level to the total range of motion. Axelsson and colleagues used the radiostereometry method to study the lumbar kinematics of fusion patients. Their results showed that, although hypermobility of the adjacent segment was seen in individual patients after a spinal fusion, it was relatively infrequent. At five-years after lumbar fusion, no significant differences between the preoperative and postoperative motions were detected in the nine fusion patients included in the study. Interindividual differences were significant, with the adjacent segments showing either unchanged, increased or decreased motion compared with the preoperative situation. Auerbach and colleagues used videofluoroscopy in five patients after circumferential lumbar fusion and noticed that the angular motion of the proximal adjacent segment was significantly more than in asymptomatic controls. However, because they did not have information on the preoperative range of motion of their fusion patients, the results of this small group of patients might reflect the interindividual variability shown by Axelsson and colleagues. In conclusion, whether fusion patients will try to maintain their preoperative lumbar range of motion postoperatively or are content with a restricted mobility is still a matter of debate.

Alterations in the lumbar muscle function probably play a significant role in the in vivo kinematics and kinetics of the spine after a fusion procedure. Damage to these muscles undoubtedly happens but, to date, this has been an understudied effect.

Adjacent Segment Intradiscal Pressure

Most in vitro biomechanical studies have demonstrated significant changes in the intradiscal pressure (IDP) of the adjacent disks after simulated fusion suggesting altered anterior column stresses. The superior adjacent segment IDP increased with flexion loading, and increased and decreased IDP has been noted with extension loading. With flexion loading, increases in IDP seem more marked after longer fusions. Specifically, an in vitro study on human cadaveric spines found the IDP to be increased by 30% after one-level fusion, and by 82% after two-level fusion. Again, changes in IDP have been shown in all unfused levels. Contrary to the previous studies, a validated FEM detected only slight changes in the IDP by rigid fixation. in vivo IDPs after lumbar fusion are difficult to measure and remain unknown.

These results must be interpreted with caution in light of the previous discussion on biomechanical testing methods used in the assessment of adjacent segment changes. If a load-controlled strategy is used there will not be substantial pressure changes in the adjacent segment intervertebral disks. However, if a displacement-controlled strategy is adopted, then the loading across the adjacent segment will increase and this will result in increased intradiscal pressures. As outlined previously, it is not known which protocol is more germane to the clinical scenario.

Adjacent Segment Facet Joint Loads

Increased facet joint forces at the level above a fusion, suggesting increased posterior column stresses, have been demonstrated by Lee and Langrana using human cadaveric specimens and by Rohlmann and colleagues in a validated FEM, especially with flexion loading. Corroborating these data with physiologic in vivo loading is difficult. The previous discussion regarding the link between the biomechanical methodology used and the observed load changes at the adjacent segment applies here for the facet joint contact forces.

Only gold members can continue reading. Log In or Register to continue