- Email: [email protected]
Surgical management of cervical myelopathy dealing with the cervical–thoracic junction
The Spine Journal 6 (2006) 268S–273S
Surgical management of cervical myelopathy dealing with the cervical–thoracic junction Samir Lapsiwala, MD, Edward Benzel, MD* Cleveland Clinic Spine Institute, The Cleveland Clinic Foundation, 9500 Euclid Avenue, S80, Cleveland, OH 44195, USA
Abstract
BACKGROUND CONTEXT: The treatment of compressive cervical myelopathy is, in general, a surgical endeavor. Surgery involves decompression, often with an accompanying fusion with stabilization. The length of the fusion can vary and the decision regarding length of fusion is not always clear. PURPOSE: This study explores the fundamental principles regarding the length of fusion at the cervicothoracic junction. STUDY DESIGN/SETTING: A review of the literature regarding the anatomy and biomechanics of the cervicothoracic region is provided. Surgical approaches and indications for cervicothoracic junction region fusions are discussed. Fundamental guidelines for the decision-making process are provided. CONCLUSION: The cervicothoracic region is a biomechanically complex region. Although there is little biomechanical data indicating the appropriate length of fusion, several fundamental guidelines may be followed to reduce the incidence of construct failure. A long fusion should not end at an apical vertebra nor at the cervicothoracic junction. Long cervical fusions should be extended to traverse the cervicothoracic junction to a neutral vertebra. Ó 2006 Elsevier Inc. All rights reserved.
Keywords:
Cervical myelopathy; Fusion; Cervicothoracic junction; Length of fusion; Cervicothoracic fusion; Cervical laminectomy
1. Introduction The management of cervical myelopathy requires decompression of the spinal cord either ventrally, dorsally, or both, depending on the pathology, geometry, and spinal alignment. Cervical laminectomy is valuable for the management of congenital or acquired cervical spinal stenosis at multiple levels. The decision to fuse or not depends on factors such as the patient’s age, preoperative sagittal alignment, underlying diagnosis, as well as the length of the decompression required. The true incidence of postlaminectomy kyphosis is difficult to ascertain from the literature because of the heterogeneous patient population and inconsistent reporting. When multilevel anterior procedures are performed, grafting and plating is essential. Fusion rates
are increased and the incidence of iatrogenic deformity is decreased [1–4]. Surgery at or near the cervicothoracic junction poses a particular challenge owing to the anatomical challenges associated with the surgical approach and the complex nature of the regional biomechanics. There exists very little literature assessing the appropriate length of fusion in the cervical spine. Length of fusion depends on factors such as age of the patient, underlying pathology, length of laminectomy, and quality of the bone. There are some general guidelines, however, that may reduce the incidence of construct failure. These are based on sound anatomical and biomechanical principles.
2. Anatomy FDA device/drug status: not applicable. * Corresponding author. Chairman, Cleveland Clinic Spine Institute, Vice Chairman, Department of Neurosurgery, The Cleveland Clinic Foundation, 9500 Euclid Avenue, S80, Cleveland, OH 44195. Tel.: (216) 4447381; fax: (216) 445-9999. E-mail address: [email protected] (E. Benzel) 1529-9430/06/$ – see front matter Ó 2006 Elsevier Inc. All rights reserved. doi:10.1016/j.spinee.2006.05.008
The vertebrae of the subaxial cervical spine are fairly uniform and are aligned in a slightly lordotic posture. The components of the cervical vertebrae include the body, superior and inferior articular processes, pedicles, lamina, and a spinous process. The vertebral bodies are the axial load-bearing elements of the spine. The pedicle of the
S. Lapsiwala and E. Benzel / The Spine Journal 6 (2006) 268S–273S
subaxial cervical spine connects the body with lateral masses. They are small and medially oriented. The lateral masses of the subaxial cervical spine consist of superior and inferior articulating processes that form the bony confines of the facet joint. The orientation of the facet joints is in the coronal plane. This limits spinal movement in extension. The cervical laminae are thin, and the spinous processes of the midcervical spine are small and often bifid. The anatomy of the thoracic spine contrasts significantly with that of the cervical spine in large part because of complex osteo-ligamentous articulation with the thoracic rib cage. An understanding of the complex relationship between transverse process, the pedicle, the neuroforamen, the rib head articulation with the vertebral body, and disc is essential for proper and safe surgery in this region. Physiologic kyphosis of the thoracic spine results from a relatively greater height of the dorsal vertebral wall compared with the ventral vertebral wall. Pedicle orientation and dimensions vary significantly over the levels of the thoracic spine [5,6]. Facet orientation in the thoracic spine is primarily in the coronal plane, which permits rotational motion. The lamina increases in width and thickness from the upper to the lower thoracic spine. As is evident from a perusal of Table 1, the range of motion in the spine depends on the level [7]. The anatomy and biomechanics of the cervicothoracic junction are distinctive. There is a transition from cervical lordosis to kyphosis in the thoracic region [5,6]. The cervical spine has very mobile vertebral elements, whereas the thoracic spine is characterized by its stabilizing facet and rib architecture (Table 1). This abrupt change from mobility to stability predisposes this region of the spine to trauma and degeneration. Special care must be taken when operating near the cervicothoracic junction. A higher incidence of postoperative kyphotic deformity exists in this region. Inoue et al. reported 36 patients whose spinal cord tumors were resected either via laminectomy, laminoplasty, or Table 1 Range of motion of intact cervicothoracic spine in three modes of motion Combined flexion/extension
One side lateral bending
One side axial rotation
Degrees C0–C1 C1–C2 C2–C3 C3–C4 C4–C5 C5–C6 C6–C7 C7–T1 T1–T2 T2–T3 T3–T4
25 20 10 15 20 20 15 10 5 5 5
5 5 10 12 12 10 10 5 5 5 5
4 40 3 7 7 7 5 1 9 8 8
Adapted from White and Punjabi [7]. Note how upper thoracic spine provides some motion between its levels.
269S
hemilaminectomy [8]. Patients who underwent C7 laminectomy developed kyphosis localized at the cervicothoracic junction and marked compensatory lordosis of the cervical region, whereas patients who underwent laminoplasty developed significantly less deformity. Steinmetz et al. retrospectively reviewed all operations involving the cervicothoracic junction in the Department of Neurosurgery at the Cleveland Clinic Foundation for a 5-year period [9]. Out of 593 total cases there were 14 failures. An extensive search for factors associated with failure established two surgical procedures with poor outcome. These include uninstrumented laminectomy and ventral multilevel corpectomies across the cervicothoracic junction. Laminectomy across the cervicothoracic junction without instrumentation was strongly associated with failure (38% of cases).
3. Surgical approach Multiple procedures allow access to the cervicothoracic junction. The nature of the pathology guides the choice of surgical approach. The ventral approach to the cervical spine has been effectively used for 50 years [10,11]. This approach provides direct access for spinal cord and nerve root decompression secondary to herniated intervertebral discs, dorsal vertebral body osteophytes, ossified posterior longitudinal ligament, and uncovertebral joint hypertrophy. Another important benefit of the ventral approach is the ability to correct kyphotic deformity, through discectomy or corpectomy [9]. Ventral approaches afford significant opportunities for deformity correction that are not possible with dorsal approaches. Ventral decompression options include discectomy and corpectomy, either single or multiple levels, or a combination of both. Arthrodesis can be accomplished by using autograft or allograft bone, or a combination of cages. Typically, cages are packed with autograft or allograft bone to facilitate fusion. All of these options can be used with or without a ventral cervical plating system. The primary advantage of a ventral approach relates to the facilitation of ventral pathology resection. Often, however, the approach is difficult, especially in the face of multiple level pathology, reoperation, or cervicothoracic junction involvement. Finally, a ventral approach nearly always obligates a simultaneous fusion procedure. Generally, dorsal approaches are reserved for patients with multilevel, predominantly dorsal or circumferential compression, in the presence of a straight or lordotic cervical alignment. This surgical option appears to be surgeon-dependent more than scientifically based. Absolute contraindication to laminectomy is preoperative kyphosis. Laminectomy may be considered in cases of cervical lordosis and often in patients with a straightened spine. Laminoplasty is an alternate to multilevel laminectomy. The goal of a laminoplasty procedure is the expansion of the spinal canal cross-sectional dimensions, a sparing of spinal stability, and the preservation of lamina integrity.
270S
S. Lapsiwala and E. Benzel / The Spine Journal 6 (2006) 268S–273S
Laminoplasty is a motion-preserving technique that has been associated with good clinical outcomes [12,13]. The spinal canal is expanded via either a unilateral or bilateral hinge type elevation of the laminae. Preservation of dorsal spinal structures facilitates reinsertion of the nuchal muscles at the time of closure. Laminoplasty has also been purported to diminish the incidence of postlaminectomy scaring and resultant neural compression. Hirabayashi and associates did not observe postoperative kyphosis of the cervical spine after open-door laminoplasty [12].
4. To fuse or not to fuse Successful surgical planning requires a careful consideration of surgical approach, length of decompression, the use and type of fusion, the use and type of instrumentation, the length of fusion, and the use of postoperative bracing. White and Panjabi described four general indication categories for spinal fusion: 1) to restore clinical stability to a spine in which the structural integrity has been compromised; 2) to maintain alignment after deformity correction; 3) to prevent deformity prevention; and 4) to alleviate pain [7]. Cervical spinal instrumentation is applied with autograft, allograft, or both in these scenarios. Rarely, instrumentation may be used alone as a sole means of spine
stabilization, both short and long term. Such a strategy may be employed in selected metastatic spine tumor cases. In the cervical spine, the decision to fuse or not is guided by many factors, including: patient age, surgical approach, preoperative sagittal alignment, underlying diagnosis, and length of decompression. Postoperative spinal deformities can result from inadequate consideration of overall spinal stability, especially in young patients and those with minimal degenerative changes who undergo laminectomy at multiple levels. Patients with extensive degenerative changes often have less overall spinal mobility; as such, degenerative changes can in fact enhance spine stability. Children have the greatest incidence of postlaminectomy kyphosis [14–18]. Tachdjian and Matson reported a 40% incidence of cervical kyphosis in 115 children undergoing multilevel laminectomy [18]. Certain disease processes increase the incidence of postlaminectomy kyphosis. Patients with neurological disease, such as intrinsic tumors or syrinx of the spinal cord, often develop spinal deformity. This may be the result of asymmetric muscle activity caused by altered neurological function or adjuvant treatment, such as postoperative radiation therapy for a primary pathology. In 2005, de Jonge et al. reported a retrospective study of 76 children with malignant spine tumors treated with laminectomy or laminoplasty and/or radiation therapy [17]. Sixty-seven (88%) developed
Fig. 1. An apical vertebra occurs at the horizon or apex of a curve, be it in the sagittal (A) or coronal plane (B). It is associated with adjacent disc interspaces that have the greatest segmental angulation (a) of all interspaces in the curve, as depicted. (Reprinted by permission from Benzel [23]).
S. Lapsiwala and E. Benzel / The Spine Journal 6 (2006) 268S–273S
postlaminectomy/postradiation spinal deformity. In 46 (60%) cases the deformity was treated surgically. Radiation therapy is particularly prone to affect the growing spine, leading to asymmetric vertebral growth and kyphosis or scoliosis. In both of these scenarios, stabilization and fusion procedures should be considered. A careful evaluation of cervical sagittal alignment is imperative. Without such, a postoperative kyphosis may result. The normal sagittal weight-bearing axis lies dorsal to the vertebral bodies of C2–C7. Maintaining normal lordotic sagittal alignment minimizes the demand on the dorsal cervical musculature to maintain balance. With a loss of normal sagittal cervical alignment, the weight-bearing axis shifts ventrally. This shift places the cervical musculature at a significant mechanical disadvantage, thus requiring constant muscular contraction to maintain an upright head posture. Eventually, muscles fatigue. This may lead to pain and possibly kyphosis. Pal and Sherk tested the concept of the three-column cervical spine via load transmission through each column [19]. They applied loads to the superior articular surfaces of the axis and recorded loads from each column separately at C6. They found that 36% of the total load applied rostrally is transmitted through the ventral column formed by vertebral bodies and intervertebral discs, and 32% each through the two dorsal cervical columns formed by the articular processes. Because the dorsal elements are responsible for the majority of load transmission, a significant loss of dorsal element integrity can result in instability. Acute stability of the cervical spine was tested after laminectomy and progressive staged foraminotomies in an in vitro model [20]. Testing showed that dorsal strain did not differ among the intact specimens and those that had been treated with laminectomy and those that had been treated with a 25% facetectomy. A 50% facetectomy resulted in a 2.5% increase in dorsal strain, and a 75% or 100% facetectomy, in a 25% increase in dorsal strain (compared with the intact specimen). In short, a fusion should be considered when an aggressive laminectomy and foraminotomy is performed. A compelling advantage of augmenting a fusion with a spinal implant is the immediate acquisition of postoperative stability (while osseous fusion is established). Such immediate stability protects neural structures from trauma (eg, by preventing graft dislodgement) and reduces the incidence of spinal deformity. Instrumentation also obviates or significantly reduces a need for an external brace. Instrumentation increases the probability of attaining successful osseous fusion. Many studies have demonstrated increased fusion rates in single and multilevel ventral discectomy procedures when an adjunctive plating system was used [2,21,22]. The disadvantages of instrumentation include complications due to malposition of the implant, improper construct design, and implant failure or construct failure. If the instrumentation is excessively rigid, it may reduce the
271S
incidence of fusion by stress shielding. For example, static plates are load-bearing devices and, as such, off-load the graft by bearing a significant portion of the load. Fusion at the rostral and caudal graft–host interfaces relies, to one degree or another, on the graft bearing at least a fraction of the load. If the implant bears the entire load, a reduced fusion rate is to be expected. Dynamic plate systems facilitate some settling, thus sharing the load between the implant and graft. This provides a more natural loading of the spine during loading. Stress shielding of the graft–host interface is, thus, minimized.
5. Length of fusion The literature regarding the optimal length of fusion is sparse. Nevertheless, several basic principles must be heeded in designing a construct. A careful study of preoperative imaging should assist in determining the length of fusion. Identifying both focal and gradual curves, as well as the apical and neutral vertebrae in sagittal and coronal planes are extremely important [23]. The apical vertebra is defined as that vertebra that is associated with the greatest segmental angulation at both its rostral and caudal disc interspaces, compared with all other disc interspace in the curve. Generally it is located in the midportion of the curve and at the ‘‘horizon’’ of the curve (Fig. 1). A neutral vertebra is one that is associated with little or no angulations at its rostral and caudal disc spaces. It is the vertebra that is located between curves (Fig. 2). The determination of the apical and neutral vertebra should be assessed in both the
Fig. 2. The neutral vertebra are located between curves, be it in the sagittal (A) or coronal plane (B). There is little or no angle between its rostral and caudal disc interspaces (b). (Reprinted by permission from Benzel [23]).
272S
S. Lapsiwala and E. Benzel / The Spine Journal 6 (2006) 268S–273S
coronal and sagittal planes. The apical vertebra is the most critical vertebra in a curve. If a long moment arm construct is extended to an apical vertebra in either the sagittal or coronal plane, a progressive deformity may result. Therefore, as a general rule, a long construct should usually not terminate at or near an apical vertebra [23]. In this case the construct should probably be lengthened to extend to a neutral vertebra. The lower cervical and upper thoracic spine is indeed a junctional region. Here, the cervical lordosis transitions into a thoracic kyphosis. In addition, a transition from smaller cervical vertebrae to larger thoracic vertebrae is observed. Despite the fact that junctional levels, such as the cervicothoracic junction, usually encompass a neutral vertebra in the sagittal plane, junctional regions are prone to deformity and deformity progression if the construct ends at these levels. The disc spaces adjacent to these neutral levels are not parallel to the ground in standing position (Fig. 3). These angulations result in the application of translation and angular stresses in the cervicothoracic junction.
Fusions that terminate in this region are prone to angular and translational deformation (Fig. 4) [23]. Fusion and instrumentation procedures that extend beyond this region, particularly if a laminectomy or foraminotomy has been performed, are optimal. For example, when C3–C7 multilevel laminectomy and fusion is performed, one may consider extending the instrumentation and fusion to the upper thoracic spine. The upper thoracic spine does indeed provide some motion in flexion/extension, lateral bending, and axial rotation (Table 1). One must also consider other factors such as loss of motion segments, complications involving extra levels of fusion, and cost of implants in determining the length of fusion. The nature and character of the pathology play a role in determining the length of fusion. For instance, patients with ankylosing spondylitis often require a longer construct [24]. Also, a consideration of bony quality is very important. Patients with rheumatoid arthritis or severe osteoporosis require the employment of additional points of fixation to avoid construct failure. Multiple studies have demonstrated a correlation between the bone mineral density of the vertebral body and the pullout strength of the instrumentations [25].
6. Conclusion The cervicothoracic region is a biomechanically complex region. In patients with cervical myelopathy, and in whom extensive decompression is performed, fusion with or without instrumentation is often necessary. Even though there exists no biomechanical data supporting a specific length of fusion, some basic guidelines should be heeded to reduce the incidence of construct failure. A long fusion
Fig. 3. The cervicothoracic region is prone to further deformation if an implant is terminated here, because of the transitional nature of this region and the fact that the disc interspaces are not parallel to the ground while standing. Both transitional and angular forces are applied and fusions that end here are prone to angular deformation. (Reprinted by permission from Benzel [23]).
Fig. 4. A long implant terminating at the cervicothoracic junction (A) places significant stress at the junction, causing junctional instability (B). (Reprinted by permission from Benzel [23]).
S. Lapsiwala and E. Benzel / The Spine Journal 6 (2006) 268S–273S
perhaps should not end at the apical vertebra or the cervicothoracic junction. Instead, the fusion should be extended beyond the cervicothoracic junction to the region of a neutral vertebra.
Acknowledgment Author EB acknowledges a financial relationship (research support from and consultant with DePuy Acromed, Mazor, Nuvasive) that may indirectly relate to the subject of this research. References [1] Wang JC, McDonough PW, Endow KK, Delamarter RB. Increased fusion rates with cervical plating for two-level anterior cervical discectomy and fusion. Spine 2000;25:41–5. [2] Wang JC, McDonough PW, Kanim LE, Endow KK, Delamarter RB. Increased fusion rates with cervical plating for three-level anterior cervical discectomy and fusion. Spine 2001;26:643–6; discussion 646–7. [3] Bose B. Anterior cervical instrumentation enhances fusion rates in multilevel reconstruction in smokers. J Spinal Disord 2001;14:3–9. [4] Geer CP, Papadopoulos SM. The argument for single-level anterior cervical discectomy and fusion with anterior plate fixation. Clin Neurosurg 1999;45:25–9; discussion 21. [5] Stanescu S, Ebraheim NA, Yeasting R, Bailey AS, Jackson WT. Morphometric evaluation of the cervico-thoracic junction: practical considerations for posterior fixation of the spine. Spine 1994;19: 2082–8. [6] An HS, Wise JJ, Xu R. Anatomy of the cervicothoracic junction: a study of cadaveric dissection, cryomicrotomy, and magnetic resonance imaging. J Spinal Disord 1999;12:519–25. [7] White AA, Panjabi MM. Clinical biomechanics of the spine. 2nd ed. Philadelphia: Lippincott Williams & Wilkins, 1990:45. [8] Inoue A, Ikata T, Katoh S. Spinal deformity following surgery for spinal cord tumors and tumorous lesions: analysis based on an assessment of the spinal functional curve. Spinal Cord 1996;34:536–42. [9] Steinmetz MP, Kager CD, Benzel EC. Ventral correction of postsurgical cervical kyphosis. J Neurosurg 2003;98(1 Suppl):1–7.
273S
[10] Robinson RA, Smith GW. Anterolateral cervical disc removal and interbody fusion for cervical disc syndrome. Bull John Hopkins Hosp 1955;96:223–4. [11] Cloward RB. The anterior approach for removal of ruptured cervical disks. J Neurosurg 1958;15:602–17. [12] Hirabayashi K, Watanabe K, Wakano K, Suzuki N, Satomi K, Ishii Y. Expansive open-door laminoplasty for cervical spinal stenotic myelopathy. Spine 1983;8:693–9. [13] Kawai S, Sunago K, Doi K, Saika M, Taguchi T. Cervical laminoplasty (Hattori’s method): procedure and follow-up results. Spine 1988;13:1245–50. [14] Albert TJ, Vacarro A. Postlaminectomy kyphosis. Spine 1998;23: 2738–45. [15] Lonstein JE. Post-laminectomy kyphosis. Clin Orthop Relat Res 1977;128:93–100. [16] Cattell HS, Clark GL. Cervical kyphosis and instability following multiple laminectomies in children. J Bone Joint Surg Am 1967; 49:713–20. [17] de Jonge T, Slullitel H, Dubousset J, Miladi L, Wicart P, Illes T. Lateonset spinal deformities in children treated by laminectomy and radiation therapy for malignant tumours. Eur Spine J 2005;14:765–71. [18] Tachdjian MO, Matson DD. Orthopaedic aspects of intraspinal tumors in infants and children. J Bone Joint Surg Am 1965;47: 223–48. [19] Pal GP, Sherk HH. The vertical stability of the cervical spine. Spine 1988;13:447–9. [20] Zdeblick TA, Zou D, Warden KE, McCabe R, Kunz D, Vanderby R. Cervical stability after foraminotomy: a biomechanical in vitro analysis. J Bone Joint Surg Am 1992;74:22–7. [21] Kaiser MG, Haid RW Jr, Subach BR, Barnes B, Rodts GE Jr. Anterior cervical plating enhances arthrodesis after discectomy and fusion with cortical allograft. Neurosurgery 2002;50:229–36; discussion 236–8. [22] Wang JC, McDonough PW, Endow K, Kanim LE, Delamarter RB. The effect of cervical plating on single-level anterior cervical discectomy and fusion. J Spinal Disord 1999;12:467–71. [23] Benzel EC. Biomechanics of spine stabilization. 2nd ed. New York: Thieme, 2001:375–410. [24] Taggard DA, Traynelis VC. Management of cervical spinal fractures in ankylosing spondylitis with posterior fixation. Spine 2000;25: 2035–9. [25] Yamagata M, Kitahara H, Minami S, et al. Mechanical stability of the pedicle screw fixation systems for the lumbar spine. Spine 1992;17(3 Suppl):S51–4.
Surgical management of cervical myelopathy dealing with the cervical–thoracic junction Samir Lapsiwala, MD, Edward Benzel, MD* Cleveland Clinic Spine Institute, The Cleveland Clinic Foundation, 9500 Euclid Avenue, S80, Cleveland, OH 44195, USA
Abstract
BACKGROUND CONTEXT: The treatment of compressive cervical myelopathy is, in general, a surgical endeavor. Surgery involves decompression, often with an accompanying fusion with stabilization. The length of the fusion can vary and the decision regarding length of fusion is not always clear. PURPOSE: This study explores the fundamental principles regarding the length of fusion at the cervicothoracic junction. STUDY DESIGN/SETTING: A review of the literature regarding the anatomy and biomechanics of the cervicothoracic region is provided. Surgical approaches and indications for cervicothoracic junction region fusions are discussed. Fundamental guidelines for the decision-making process are provided. CONCLUSION: The cervicothoracic region is a biomechanically complex region. Although there is little biomechanical data indicating the appropriate length of fusion, several fundamental guidelines may be followed to reduce the incidence of construct failure. A long fusion should not end at an apical vertebra nor at the cervicothoracic junction. Long cervical fusions should be extended to traverse the cervicothoracic junction to a neutral vertebra. Ó 2006 Elsevier Inc. All rights reserved.
Keywords:
Cervical myelopathy; Fusion; Cervicothoracic junction; Length of fusion; Cervicothoracic fusion; Cervical laminectomy
1. Introduction The management of cervical myelopathy requires decompression of the spinal cord either ventrally, dorsally, or both, depending on the pathology, geometry, and spinal alignment. Cervical laminectomy is valuable for the management of congenital or acquired cervical spinal stenosis at multiple levels. The decision to fuse or not depends on factors such as the patient’s age, preoperative sagittal alignment, underlying diagnosis, as well as the length of the decompression required. The true incidence of postlaminectomy kyphosis is difficult to ascertain from the literature because of the heterogeneous patient population and inconsistent reporting. When multilevel anterior procedures are performed, grafting and plating is essential. Fusion rates
are increased and the incidence of iatrogenic deformity is decreased [1–4]. Surgery at or near the cervicothoracic junction poses a particular challenge owing to the anatomical challenges associated with the surgical approach and the complex nature of the regional biomechanics. There exists very little literature assessing the appropriate length of fusion in the cervical spine. Length of fusion depends on factors such as age of the patient, underlying pathology, length of laminectomy, and quality of the bone. There are some general guidelines, however, that may reduce the incidence of construct failure. These are based on sound anatomical and biomechanical principles.
2. Anatomy FDA device/drug status: not applicable. * Corresponding author. Chairman, Cleveland Clinic Spine Institute, Vice Chairman, Department of Neurosurgery, The Cleveland Clinic Foundation, 9500 Euclid Avenue, S80, Cleveland, OH 44195. Tel.: (216) 4447381; fax: (216) 445-9999. E-mail address: [email protected] (E. Benzel) 1529-9430/06/$ – see front matter Ó 2006 Elsevier Inc. All rights reserved. doi:10.1016/j.spinee.2006.05.008
The vertebrae of the subaxial cervical spine are fairly uniform and are aligned in a slightly lordotic posture. The components of the cervical vertebrae include the body, superior and inferior articular processes, pedicles, lamina, and a spinous process. The vertebral bodies are the axial load-bearing elements of the spine. The pedicle of the
S. Lapsiwala and E. Benzel / The Spine Journal 6 (2006) 268S–273S
subaxial cervical spine connects the body with lateral masses. They are small and medially oriented. The lateral masses of the subaxial cervical spine consist of superior and inferior articulating processes that form the bony confines of the facet joint. The orientation of the facet joints is in the coronal plane. This limits spinal movement in extension. The cervical laminae are thin, and the spinous processes of the midcervical spine are small and often bifid. The anatomy of the thoracic spine contrasts significantly with that of the cervical spine in large part because of complex osteo-ligamentous articulation with the thoracic rib cage. An understanding of the complex relationship between transverse process, the pedicle, the neuroforamen, the rib head articulation with the vertebral body, and disc is essential for proper and safe surgery in this region. Physiologic kyphosis of the thoracic spine results from a relatively greater height of the dorsal vertebral wall compared with the ventral vertebral wall. Pedicle orientation and dimensions vary significantly over the levels of the thoracic spine [5,6]. Facet orientation in the thoracic spine is primarily in the coronal plane, which permits rotational motion. The lamina increases in width and thickness from the upper to the lower thoracic spine. As is evident from a perusal of Table 1, the range of motion in the spine depends on the level [7]. The anatomy and biomechanics of the cervicothoracic junction are distinctive. There is a transition from cervical lordosis to kyphosis in the thoracic region [5,6]. The cervical spine has very mobile vertebral elements, whereas the thoracic spine is characterized by its stabilizing facet and rib architecture (Table 1). This abrupt change from mobility to stability predisposes this region of the spine to trauma and degeneration. Special care must be taken when operating near the cervicothoracic junction. A higher incidence of postoperative kyphotic deformity exists in this region. Inoue et al. reported 36 patients whose spinal cord tumors were resected either via laminectomy, laminoplasty, or Table 1 Range of motion of intact cervicothoracic spine in three modes of motion Combined flexion/extension
One side lateral bending
One side axial rotation
Degrees C0–C1 C1–C2 C2–C3 C3–C4 C4–C5 C5–C6 C6–C7 C7–T1 T1–T2 T2–T3 T3–T4
25 20 10 15 20 20 15 10 5 5 5
5 5 10 12 12 10 10 5 5 5 5
4 40 3 7 7 7 5 1 9 8 8
Adapted from White and Punjabi [7]. Note how upper thoracic spine provides some motion between its levels.
269S
hemilaminectomy [8]. Patients who underwent C7 laminectomy developed kyphosis localized at the cervicothoracic junction and marked compensatory lordosis of the cervical region, whereas patients who underwent laminoplasty developed significantly less deformity. Steinmetz et al. retrospectively reviewed all operations involving the cervicothoracic junction in the Department of Neurosurgery at the Cleveland Clinic Foundation for a 5-year period [9]. Out of 593 total cases there were 14 failures. An extensive search for factors associated with failure established two surgical procedures with poor outcome. These include uninstrumented laminectomy and ventral multilevel corpectomies across the cervicothoracic junction. Laminectomy across the cervicothoracic junction without instrumentation was strongly associated with failure (38% of cases).
3. Surgical approach Multiple procedures allow access to the cervicothoracic junction. The nature of the pathology guides the choice of surgical approach. The ventral approach to the cervical spine has been effectively used for 50 years [10,11]. This approach provides direct access for spinal cord and nerve root decompression secondary to herniated intervertebral discs, dorsal vertebral body osteophytes, ossified posterior longitudinal ligament, and uncovertebral joint hypertrophy. Another important benefit of the ventral approach is the ability to correct kyphotic deformity, through discectomy or corpectomy [9]. Ventral approaches afford significant opportunities for deformity correction that are not possible with dorsal approaches. Ventral decompression options include discectomy and corpectomy, either single or multiple levels, or a combination of both. Arthrodesis can be accomplished by using autograft or allograft bone, or a combination of cages. Typically, cages are packed with autograft or allograft bone to facilitate fusion. All of these options can be used with or without a ventral cervical plating system. The primary advantage of a ventral approach relates to the facilitation of ventral pathology resection. Often, however, the approach is difficult, especially in the face of multiple level pathology, reoperation, or cervicothoracic junction involvement. Finally, a ventral approach nearly always obligates a simultaneous fusion procedure. Generally, dorsal approaches are reserved for patients with multilevel, predominantly dorsal or circumferential compression, in the presence of a straight or lordotic cervical alignment. This surgical option appears to be surgeon-dependent more than scientifically based. Absolute contraindication to laminectomy is preoperative kyphosis. Laminectomy may be considered in cases of cervical lordosis and often in patients with a straightened spine. Laminoplasty is an alternate to multilevel laminectomy. The goal of a laminoplasty procedure is the expansion of the spinal canal cross-sectional dimensions, a sparing of spinal stability, and the preservation of lamina integrity.
270S
S. Lapsiwala and E. Benzel / The Spine Journal 6 (2006) 268S–273S
Laminoplasty is a motion-preserving technique that has been associated with good clinical outcomes [12,13]. The spinal canal is expanded via either a unilateral or bilateral hinge type elevation of the laminae. Preservation of dorsal spinal structures facilitates reinsertion of the nuchal muscles at the time of closure. Laminoplasty has also been purported to diminish the incidence of postlaminectomy scaring and resultant neural compression. Hirabayashi and associates did not observe postoperative kyphosis of the cervical spine after open-door laminoplasty [12].
4. To fuse or not to fuse Successful surgical planning requires a careful consideration of surgical approach, length of decompression, the use and type of fusion, the use and type of instrumentation, the length of fusion, and the use of postoperative bracing. White and Panjabi described four general indication categories for spinal fusion: 1) to restore clinical stability to a spine in which the structural integrity has been compromised; 2) to maintain alignment after deformity correction; 3) to prevent deformity prevention; and 4) to alleviate pain [7]. Cervical spinal instrumentation is applied with autograft, allograft, or both in these scenarios. Rarely, instrumentation may be used alone as a sole means of spine
stabilization, both short and long term. Such a strategy may be employed in selected metastatic spine tumor cases. In the cervical spine, the decision to fuse or not is guided by many factors, including: patient age, surgical approach, preoperative sagittal alignment, underlying diagnosis, and length of decompression. Postoperative spinal deformities can result from inadequate consideration of overall spinal stability, especially in young patients and those with minimal degenerative changes who undergo laminectomy at multiple levels. Patients with extensive degenerative changes often have less overall spinal mobility; as such, degenerative changes can in fact enhance spine stability. Children have the greatest incidence of postlaminectomy kyphosis [14–18]. Tachdjian and Matson reported a 40% incidence of cervical kyphosis in 115 children undergoing multilevel laminectomy [18]. Certain disease processes increase the incidence of postlaminectomy kyphosis. Patients with neurological disease, such as intrinsic tumors or syrinx of the spinal cord, often develop spinal deformity. This may be the result of asymmetric muscle activity caused by altered neurological function or adjuvant treatment, such as postoperative radiation therapy for a primary pathology. In 2005, de Jonge et al. reported a retrospective study of 76 children with malignant spine tumors treated with laminectomy or laminoplasty and/or radiation therapy [17]. Sixty-seven (88%) developed
Fig. 1. An apical vertebra occurs at the horizon or apex of a curve, be it in the sagittal (A) or coronal plane (B). It is associated with adjacent disc interspaces that have the greatest segmental angulation (a) of all interspaces in the curve, as depicted. (Reprinted by permission from Benzel [23]).
S. Lapsiwala and E. Benzel / The Spine Journal 6 (2006) 268S–273S
postlaminectomy/postradiation spinal deformity. In 46 (60%) cases the deformity was treated surgically. Radiation therapy is particularly prone to affect the growing spine, leading to asymmetric vertebral growth and kyphosis or scoliosis. In both of these scenarios, stabilization and fusion procedures should be considered. A careful evaluation of cervical sagittal alignment is imperative. Without such, a postoperative kyphosis may result. The normal sagittal weight-bearing axis lies dorsal to the vertebral bodies of C2–C7. Maintaining normal lordotic sagittal alignment minimizes the demand on the dorsal cervical musculature to maintain balance. With a loss of normal sagittal cervical alignment, the weight-bearing axis shifts ventrally. This shift places the cervical musculature at a significant mechanical disadvantage, thus requiring constant muscular contraction to maintain an upright head posture. Eventually, muscles fatigue. This may lead to pain and possibly kyphosis. Pal and Sherk tested the concept of the three-column cervical spine via load transmission through each column [19]. They applied loads to the superior articular surfaces of the axis and recorded loads from each column separately at C6. They found that 36% of the total load applied rostrally is transmitted through the ventral column formed by vertebral bodies and intervertebral discs, and 32% each through the two dorsal cervical columns formed by the articular processes. Because the dorsal elements are responsible for the majority of load transmission, a significant loss of dorsal element integrity can result in instability. Acute stability of the cervical spine was tested after laminectomy and progressive staged foraminotomies in an in vitro model [20]. Testing showed that dorsal strain did not differ among the intact specimens and those that had been treated with laminectomy and those that had been treated with a 25% facetectomy. A 50% facetectomy resulted in a 2.5% increase in dorsal strain, and a 75% or 100% facetectomy, in a 25% increase in dorsal strain (compared with the intact specimen). In short, a fusion should be considered when an aggressive laminectomy and foraminotomy is performed. A compelling advantage of augmenting a fusion with a spinal implant is the immediate acquisition of postoperative stability (while osseous fusion is established). Such immediate stability protects neural structures from trauma (eg, by preventing graft dislodgement) and reduces the incidence of spinal deformity. Instrumentation also obviates or significantly reduces a need for an external brace. Instrumentation increases the probability of attaining successful osseous fusion. Many studies have demonstrated increased fusion rates in single and multilevel ventral discectomy procedures when an adjunctive plating system was used [2,21,22]. The disadvantages of instrumentation include complications due to malposition of the implant, improper construct design, and implant failure or construct failure. If the instrumentation is excessively rigid, it may reduce the
271S
incidence of fusion by stress shielding. For example, static plates are load-bearing devices and, as such, off-load the graft by bearing a significant portion of the load. Fusion at the rostral and caudal graft–host interfaces relies, to one degree or another, on the graft bearing at least a fraction of the load. If the implant bears the entire load, a reduced fusion rate is to be expected. Dynamic plate systems facilitate some settling, thus sharing the load between the implant and graft. This provides a more natural loading of the spine during loading. Stress shielding of the graft–host interface is, thus, minimized.
5. Length of fusion The literature regarding the optimal length of fusion is sparse. Nevertheless, several basic principles must be heeded in designing a construct. A careful study of preoperative imaging should assist in determining the length of fusion. Identifying both focal and gradual curves, as well as the apical and neutral vertebrae in sagittal and coronal planes are extremely important [23]. The apical vertebra is defined as that vertebra that is associated with the greatest segmental angulation at both its rostral and caudal disc interspaces, compared with all other disc interspace in the curve. Generally it is located in the midportion of the curve and at the ‘‘horizon’’ of the curve (Fig. 1). A neutral vertebra is one that is associated with little or no angulations at its rostral and caudal disc spaces. It is the vertebra that is located between curves (Fig. 2). The determination of the apical and neutral vertebra should be assessed in both the
Fig. 2. The neutral vertebra are located between curves, be it in the sagittal (A) or coronal plane (B). There is little or no angle between its rostral and caudal disc interspaces (b). (Reprinted by permission from Benzel [23]).
272S
S. Lapsiwala and E. Benzel / The Spine Journal 6 (2006) 268S–273S
coronal and sagittal planes. The apical vertebra is the most critical vertebra in a curve. If a long moment arm construct is extended to an apical vertebra in either the sagittal or coronal plane, a progressive deformity may result. Therefore, as a general rule, a long construct should usually not terminate at or near an apical vertebra [23]. In this case the construct should probably be lengthened to extend to a neutral vertebra. The lower cervical and upper thoracic spine is indeed a junctional region. Here, the cervical lordosis transitions into a thoracic kyphosis. In addition, a transition from smaller cervical vertebrae to larger thoracic vertebrae is observed. Despite the fact that junctional levels, such as the cervicothoracic junction, usually encompass a neutral vertebra in the sagittal plane, junctional regions are prone to deformity and deformity progression if the construct ends at these levels. The disc spaces adjacent to these neutral levels are not parallel to the ground in standing position (Fig. 3). These angulations result in the application of translation and angular stresses in the cervicothoracic junction.
Fusions that terminate in this region are prone to angular and translational deformation (Fig. 4) [23]. Fusion and instrumentation procedures that extend beyond this region, particularly if a laminectomy or foraminotomy has been performed, are optimal. For example, when C3–C7 multilevel laminectomy and fusion is performed, one may consider extending the instrumentation and fusion to the upper thoracic spine. The upper thoracic spine does indeed provide some motion in flexion/extension, lateral bending, and axial rotation (Table 1). One must also consider other factors such as loss of motion segments, complications involving extra levels of fusion, and cost of implants in determining the length of fusion. The nature and character of the pathology play a role in determining the length of fusion. For instance, patients with ankylosing spondylitis often require a longer construct [24]. Also, a consideration of bony quality is very important. Patients with rheumatoid arthritis or severe osteoporosis require the employment of additional points of fixation to avoid construct failure. Multiple studies have demonstrated a correlation between the bone mineral density of the vertebral body and the pullout strength of the instrumentations [25].
6. Conclusion The cervicothoracic region is a biomechanically complex region. In patients with cervical myelopathy, and in whom extensive decompression is performed, fusion with or without instrumentation is often necessary. Even though there exists no biomechanical data supporting a specific length of fusion, some basic guidelines should be heeded to reduce the incidence of construct failure. A long fusion
Fig. 3. The cervicothoracic region is prone to further deformation if an implant is terminated here, because of the transitional nature of this region and the fact that the disc interspaces are not parallel to the ground while standing. Both transitional and angular forces are applied and fusions that end here are prone to angular deformation. (Reprinted by permission from Benzel [23]).
Fig. 4. A long implant terminating at the cervicothoracic junction (A) places significant stress at the junction, causing junctional instability (B). (Reprinted by permission from Benzel [23]).
S. Lapsiwala and E. Benzel / The Spine Journal 6 (2006) 268S–273S
perhaps should not end at the apical vertebra or the cervicothoracic junction. Instead, the fusion should be extended beyond the cervicothoracic junction to the region of a neutral vertebra.
Acknowledgment Author EB acknowledges a financial relationship (research support from and consultant with DePuy Acromed, Mazor, Nuvasive) that may indirectly relate to the subject of this research. References [1] Wang JC, McDonough PW, Endow KK, Delamarter RB. Increased fusion rates with cervical plating for two-level anterior cervical discectomy and fusion. Spine 2000;25:41–5. [2] Wang JC, McDonough PW, Kanim LE, Endow KK, Delamarter RB. Increased fusion rates with cervical plating for three-level anterior cervical discectomy and fusion. Spine 2001;26:643–6; discussion 646–7. [3] Bose B. Anterior cervical instrumentation enhances fusion rates in multilevel reconstruction in smokers. J Spinal Disord 2001;14:3–9. [4] Geer CP, Papadopoulos SM. The argument for single-level anterior cervical discectomy and fusion with anterior plate fixation. Clin Neurosurg 1999;45:25–9; discussion 21. [5] Stanescu S, Ebraheim NA, Yeasting R, Bailey AS, Jackson WT. Morphometric evaluation of the cervico-thoracic junction: practical considerations for posterior fixation of the spine. Spine 1994;19: 2082–8. [6] An HS, Wise JJ, Xu R. Anatomy of the cervicothoracic junction: a study of cadaveric dissection, cryomicrotomy, and magnetic resonance imaging. J Spinal Disord 1999;12:519–25. [7] White AA, Panjabi MM. Clinical biomechanics of the spine. 2nd ed. Philadelphia: Lippincott Williams & Wilkins, 1990:45. [8] Inoue A, Ikata T, Katoh S. Spinal deformity following surgery for spinal cord tumors and tumorous lesions: analysis based on an assessment of the spinal functional curve. Spinal Cord 1996;34:536–42. [9] Steinmetz MP, Kager CD, Benzel EC. Ventral correction of postsurgical cervical kyphosis. J Neurosurg 2003;98(1 Suppl):1–7.
273S
[10] Robinson RA, Smith GW. Anterolateral cervical disc removal and interbody fusion for cervical disc syndrome. Bull John Hopkins Hosp 1955;96:223–4. [11] Cloward RB. The anterior approach for removal of ruptured cervical disks. J Neurosurg 1958;15:602–17. [12] Hirabayashi K, Watanabe K, Wakano K, Suzuki N, Satomi K, Ishii Y. Expansive open-door laminoplasty for cervical spinal stenotic myelopathy. Spine 1983;8:693–9. [13] Kawai S, Sunago K, Doi K, Saika M, Taguchi T. Cervical laminoplasty (Hattori’s method): procedure and follow-up results. Spine 1988;13:1245–50. [14] Albert TJ, Vacarro A. Postlaminectomy kyphosis. Spine 1998;23: 2738–45. [15] Lonstein JE. Post-laminectomy kyphosis. Clin Orthop Relat Res 1977;128:93–100. [16] Cattell HS, Clark GL. Cervical kyphosis and instability following multiple laminectomies in children. J Bone Joint Surg Am 1967; 49:713–20. [17] de Jonge T, Slullitel H, Dubousset J, Miladi L, Wicart P, Illes T. Lateonset spinal deformities in children treated by laminectomy and radiation therapy for malignant tumours. Eur Spine J 2005;14:765–71. [18] Tachdjian MO, Matson DD. Orthopaedic aspects of intraspinal tumors in infants and children. J Bone Joint Surg Am 1965;47: 223–48. [19] Pal GP, Sherk HH. The vertical stability of the cervical spine. Spine 1988;13:447–9. [20] Zdeblick TA, Zou D, Warden KE, McCabe R, Kunz D, Vanderby R. Cervical stability after foraminotomy: a biomechanical in vitro analysis. J Bone Joint Surg Am 1992;74:22–7. [21] Kaiser MG, Haid RW Jr, Subach BR, Barnes B, Rodts GE Jr. Anterior cervical plating enhances arthrodesis after discectomy and fusion with cortical allograft. Neurosurgery 2002;50:229–36; discussion 236–8. [22] Wang JC, McDonough PW, Endow K, Kanim LE, Delamarter RB. The effect of cervical plating on single-level anterior cervical discectomy and fusion. J Spinal Disord 1999;12:467–71. [23] Benzel EC. Biomechanics of spine stabilization. 2nd ed. New York: Thieme, 2001:375–410. [24] Taggard DA, Traynelis VC. Management of cervical spinal fractures in ankylosing spondylitis with posterior fixation. Spine 2000;25: 2035–9. [25] Yamagata M, Kitahara H, Minami S, et al. Mechanical stability of the pedicle screw fixation systems for the lumbar spine. Spine 1992;17(3 Suppl):S51–4.