31 Adjacent Segment Degeneration and Disease
The discussion regarding adjacent-segment “disease” has come to the forefront in recent years. This is primarily related to the marked and escalating interest in motion preservation and artificial disc surgery. It seems intuitive that a motion preservation strategy, such as a total disc arthroplasty (TDA), would be associated with a minimal effect on adjacent motion segments—the logic being that a minimal disruption of motion should not adversely affect adjacent levels. Theoretically, this would diminish the stress on adjacent motion segments, compared with a strategy that immobilized the index level (i.e., via fusion), and in turn would be associated with a diminished incidence of adjacent-segment degeneration and adjacent-segment disease (symptomatic degeneration).
As an aside, the difference between adjacent-segment degeneration and adjacent-segment disease should be clearly delineated. The two terms are (inappropriately) occasionally used interchangeably. To be clear, adjacent-segment degeneration is asymptomatic, whereas adjacent-segment disease is symptomatic. Despite such clinical differences, their radiographic and imaging characteristics can be identical. Magnetic resonance imaging can be used to identify adjacent-segment degeneration, but with a very high sensitivity. 1 One should, hence, be cautious when determining the presence or absence of adjacent-segment degeneration. The actual definition of the presence of adjacent-segment degeneration, it is emphasized, is relatively subjective.
What is forgotten in the discussion regarding the preservation of motion by first-generation TDAs is that although flexion–extension and lateral bending mobility may be retained, the quality of the motion is altered (see Chapter 32). First-generation TDAs (i.e., the devices most commonly employed today) are associated with minimal stiffness in flexion–extension and lateral bending, with a marked increased stiffness in axial loading, This greatly increases stress and loading at the same-level facet joints and adjacent motion segments.
If one considers the enthusiasm for TDA technologies, the difficulties associated with the definition of adjacent-segment pathology, and the myriad of clinical and anatomical variables at play, it is not surprising that the etiology of adjacent-segment degeneration and adjacent-segment disease is controversial and that the process itself is the subject of significant discussion and debate. 2 – 35 Many associated and relevant clinical factors exist and are discussed in this chapter. Although clinical studies predominate, the study of adjacent-segment pathology is not restricted to the clinical domain. Both computerized analyses and human and animal biomechanical models have been employed to further elucidate the relevant factors. 7 , 23 , 24 , 31 , 36 – 38 They are, as well, steeped in controversy.
The varying methodologies associated with both the clinical and laboratory studies employed to assess adjacent-segment degeneration and adjacent-segment disease and their associated risk factors most likely explain the controversy and the “shifting sand” nature of the enthusiasm for TDA and related technologies. The methodologies employed in this arena are often flawed. This raises questions regarding the conclusions derived. Regardless, a review of the available literature is relevant and presented here. This review is broken down into two segments. The “older,” pre-TDA literature is presented first in order to understand what has been known and the foundation upon which subsequent studies have been performed. It then becomes revealing to review the newer literature that was significantly influenced by the TDA era. Finally, the influence of spinal posture and sagittal balance, as they relate to adjacent-segment degeneration and adjacent-segment disease, is discussed. It is suggested that the latter, rather than the issue surrounding motion preservation or fusion, is the most relevant factor associated with end-fusion degenerative changes.
31.1 Historical Literature Review
As is addressed in detail in the next chapter (see Chapter 32), axial loads transferred to adjacent levels and to the facet joints at the same level are increased, not decreased, following the implantation of a first-generation TDA. This is due to the significant stiffness in axial loading associated with a metal-on-metal or a metal-on-polymer interbody spacer. Although flexion–extension and lateral bending are permitted with a TDA, they may indeed be offset by this axial-loading effect. In addition, it has been assumed that fusion alters adjacent-level mechanics to such an extent that the adjacent motion segments are significantly and adversely affected. Although end-fusion degenerative changes have been observed and studied for years, 39 the risk factors associated with adjacent-segment degeneration and adjacent-segment disease remain a point of significant discussion and debate. Therefore, this topic deserves further exploration via a historical review of the literature.
The annual incidence of de novo adjacent-segment disease varies from 1.5 to 4.5%. 40 – 42 In a compilation of three studies, Hilibrand et al corroborated the low incidence of adjacent-segment disease (1.5 to 4.5%) following anterior cervical discectomy and fusion (ACDF). 43 , 44 They in fact observed that the incidence of adjacent-segment disease did not differ from the natural unoperated history of the “disease”: “There appears to be an incidence of adjacent-segment degeneration and disease after arthrodesis that may be related to natural degeneration or the adjacent fusion.” 43
Lunsford et al also observed an annual incidence of adjacent-segment disease of 2.5% in a study examining ACD with and without fusion. In addition, the authors observed no difference between the incidence rates of adjacent-segment disease in the fusion and no-fusion groups. This observation suggests that neither ventral cervical spine intervention nor fusion significantly alters adjacent-segment mechanics to a clinically relevant extent. 45
In a particularly revealing study from the 1980s, Henderson et al demonstrated an annual incidence of adjacent-segment disease of 3% in more than 800 patients undergoing posterior laminoforaminotomy. 46 Posterior laminoforaminotomy is, strictly speaking, a motion preservation procedure. To observe a consistent 3% incidence of adjacent-segment disease following such a procedure strongly suggests that the incidence of adjacent-segment disease is related to factors other than fusion and that it occurs independently of surgery. In a somewhat controversial recent paper, Clarke et al observed an annual incidence of adjacent-segment disease significantly lower than that observed by Henderson et al. 32 Their methodology was, however, challenged by McCormick. 30 It is emphasized that the study of Clarke et al is modern, and the methodology and results influenced by the TDA era.
The consistently low incidence of adjacent-segment disease is further corroborated by the observations of Hilibrand and colleagues. Hilibrand et al published their large series of patients undergoing ACDF (409 patients) in 1999. They observed an annual incidence of adjacent-segment disease of 3% following ACDF. 47 This, again, does not differ from the incidence associated with the natural history. Furthermore, they observed that the incidence of adjacent-segment disease was less with multiple-level fusions than with single-level fusions. This counterintuitive observation truly deemphasizes and in fact negates the association between fusion and adjacent-segment disease. 48 From a biomechanical perspective, this finding is counterintuitive because of the notion that a longer fusion (and hence, moment arm) should be associated with greater stresses at adjacent motion segments. This, in turn, should be associated with a greater incidence of adjacent-segment disease—not a lesser incidence, as indeed was observed. 47
The explanations for the observation by Hilibrand et al are at least twofold. First, the majority of ACDFs are performed at the levels most prone to degeneration (C5–6 and C6–7). Hence, a two-level ACDF most likely would involve both C5–6 and C6–7. This leaves only relatively degeneration-resistant levels—in which the incidence of degenerative changes following surgery, or occurring naturally, is diminished. Second, it is also probable that fusion length (moment arm length) is not a significant factor regarding the development of adjacent-segment disease. Other factors, such as sagittal alignment and the restoration of normal posture at the index surgery level(s), may far supersede moment arm length regarding relevance. 49 , 50 Others, however, have biomechanically refuted these findings. 51
To add to the fray, Finn et al observed that the retention of a motion segment between two noncontiguous cervical fusions resulted in less adjacent-segment and intermediate-segment strain. This “skip fusion construct” study suggests that it might be appropriate to consider noncontiguous fusions instead of three-level fusions in selected situations. 52 As expected, controversy prevails. 53
31.2 Modern Literature Review
Much of what was historically shown regarding adjacent-segment degeneration following fusion has been confirmed with recent studies. 54 – 56 Ahn et al observed that age, the presence of degenerative disease and multiple-level fusions, and male gender correlated with adjacent-segment changes. 54
Initial biomechanical studies of TDA kinematics in cadaveric models supported the hypothesis that TDA retained native range of motion at the operated and adjacent levels. This is quite unlike ACDF, which decreases range of motion at the operated level and, as a consequence, increases motion and stresses at adjacent levels. 48 Early clinical studies also corroborated such findings. 25 – 27 Puttlitz et al compared ProDisc-C (Synthes, West Chester, PA) arthroplasty with the native intervertebral disc at C4–5 in a cadaveric model. Using a pure moment bending methodology, they observed no difference between TDA and the native disc. They concluded, without direct measurement, that motion at adjacent levels was replicated by TDA. 57 The conclusion regarding motion at adjacent levels is, however, not based on fact and may be overstated.
DiAngelo et al compared the Bristol cervical disc (Medtronic Sofamor Danek, Memphis, TN), which is similar to the Prestige cervical disc (Medtronic Sofamor Danek)—metal on metal; native disc; and graft with plate at C5–6 in a cadaver model. They found no difference between arthroplasty and native disc at the operated and adjacent levels, whereas they found decreased range of motion at the operated level and increased motion at adjacent levels with graft and plate. 58 Of significant note here is the fact that they did not control for sagittal alignment at the index level(s). This is a major methodologic deficiency.
Chang et al compared ProDisc-C (metal on polymer), Prestige (metal on metal), ACDF, and native disc at C6–7 in a cadaveric model. They observed that ACDF decreased range of motion at the operated level and increased range of motion at the adjacent levels. 59 Again, the conclusion regarding motion at adjacent levels was based on a study in which there was no control for sagittal alignment at the index level(s). Chang et al also observed an increased range of motion at the operated level in the arthroplasty group compared with the native spine and a decreased range of motion at adjacent levels in the arthroplasty group. 59
Dmitriev et al compared arthroplasty versus native versus allograft dowel versus allograft dowel and plate regarding index-level range of motion in all modes and adjacent-level intradiscal pressure at C5–6 in cadaver model. They observed no differences in range of motion at the index level in a comparison of arthroplasty and native spine. Adjacent-level intradiscal pressures were unchanged in arthroplasty versus native spine, whereas these were increased at the rostral adjacent level and trended toward increased at the caudal adjacent level in the arthrodesis constructs. 60 Anderson et al also observed that fusion and arthroplasty had similar kinematic effects on adjacent levels. 61
The aforementioned data, on the surface, appear convincing. However, more recent studies and analyses provide conflicting views. Harrod et al, for example, published their systematic review of adjacent-segment pathology following cervical motion–sparing procedures compared with fusion. They observed a paucity of high-quality literature in this arena. Regardless, they found no significant difference in the development of adjacent-segment degeneration or disease when they compared arthroplasty and fusion. Nunley et al confirmed these findings in a clinical assessment. 62 No conclusions regarding the superiority of either management strategy could be made. 63 Cho and Riew and Boselie et al recently provided reviews that are in agreement with that of Harrod et al and that demonstrated no difference in adjacent-segment disease between fusion and TDA. 56 , 64 Finally, Wu et al found a very low incidence of adjacent-segment disease following ACDF. 65 In addition, they observed that younger and male patients were more likely to develop adjacent-segment disease.
How does one make sense of the literature? It is very confusing and conflicted. It is influenced by the era in which the studies were performed. The latter suggests bias regarding the establishment of study design and interpretation. What is becoming increasingly evident, though, is that spinal posture and sagittal alignment are critical regarding the stresses placed on adjacent motion segments. Therefore, this chapter closes with a brief discussion of sagittal balance and posture as they relate to adjacent-segment degeneration and adjacent-segment disease.
Related posts:
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
