Acta Neurochir (Wien) (2006) 148: 943–950 DOI 10.1007/s00701-006-0805-6

Clinical Article Clinical and radiological results following cervical arthroplasty D. H. Yoon, S. Yi, H. C. Shin, K. N. Kim, and S. H. Kim Department of Neurosurgery, Yonsei University, College of Medicine, Seoul, Korea Received January 12, 2006; accepted April 25, 2006; Published online June 23, 2006 # Springer-Verlag 2006

Summary Background. This was a retrospective study of clinical and radiological results of cervical arthroplasty using the Bryan cervical disc prosthesis to evaluate the efficacy of arthroplasty in clinical applications. Methods. A total of 46 patients underwent arthroplasty of a single level using the Bryan disc prosthesis. Clinical outcome was assessed using the visual analogue scale (VAS) and the neck disability index (NDI). All patients were evaluated using preoperative and postoperative static cervical spine radiographs to compare cervical sagittal balance. Dynamic cervical spine radiographs were used to compare movement at the level of the procedure, movement at the adjacent level and movement of the whole cervical spine. Findings. With the exception of four patients with aggravated neck pain, the NDI and VAS scores decreased significantly in late follow-up evaluations. The range of movement of the whole cervical spine, the functional segmental unit, and the adjacent segments were preserved after arthroplasty. The sagittal alignment of the cervical spine showed kyphosis after surgery but restored lordosis at a later time. The postulated cause of kyphotic changes include ‘‘over-milling’’ at the dorsal endplate, inappropriate angle of disc insertion, structural absence of lordosis in the Bryan disc, removal of posterior longitudinal ligament, and pre-existing kyphosis. Conclusions. Arthroplasty using the Bryan disc appears to be safe and provided a favorable preliminary clinical and radiological outcome. Postoperative kyphosis can be prevented by understanding the biomechanical properties of the Bryan disc. Future studies will need to address the association between postoperative kyphosis, clinical outcome and adjacent segment disease. Keywords: Adjacent segment disease; Bryan disc; cervical arthroplasty; clinical outcome; kyphosis; radiological outcome.

Introduction Anterior cervical discectomy and fusion is a wellestablished, commonly performed surgical procedure for cervical spondylosis. Since its introduction in the 1950s by Robinson and Smith as well as Cloward, excel-

lent clinical results have been reported in the treatment of degenerative disorders of the cervical spine [7, 10, 20, 24]. The primary disadvantage of the procedure is that interbody fusion converts a functionally mobile, mechanically stable spinal unit into a fixed, nonfunctional unit. Analysis of the strain distribution of intervertebral discs after anterior cervical disc fusion has shown an increase in longitudinal strain, most frequently at the levels immediately adjacent to the fused segment [26]. The resulting increase in stress on adjacent discs is thought to lead to accelerated disc degeneration and mechanical instability [3]. Radiographic changes consistent with spondylosis and instability at levels above and below cervical fusions have been described by several authors [3, 4, 12, 20, 22, 31] but these changes have not always manifested with clinical symptoms. Recently, cervical arthroplasty performed with artificial cervical discs has gained attention as an alternative to traditional arthrodesis and can be used to restore and maintain mobility and function of the involved cervical spinal segments [2, 9, 15, 16]. The theoretical advantages of disc arthroplasty include: maintenance of a range of movement, avoidance of adjacent segment degeneration, reconstitution of disc height and spinal alignment, and greater maintenance of manoeuverability. Furthermore, the procedure shows decreased surgical morbidity, avoidance of complications from instrumentation or postoperative immobilization, and allows an earlier return to the previous level of function [2]. We describe the clinical and radiographic outcomes in a cohort of patients retrospectively enrolled who underwent cervical disc arthroplasty. We limited our study to

944

those patients who had single-level arthroplasty with the Bryan cervical disc system (Medtronic Sofamor Danek, Memphis, TN, USA).

D. H. Yoon et al. neoplastic disease, severe kyphosis and osteoporosis. Surgical indications are discussed precisely in the later.

Radiological analysis

Patients and methods Patient population We performed 46 arthroplasties using the Bryan disc from 2003 to 2005. The study consisted of 24 men and 22 women whose mean age was 42.3 years (range 26–58). The mean clinical follow up duration was 11.8 months (range 2.9–19.5). Single-level arthroplasty was performed in 34 patients and 12 patients underwent single-level arthroplasty with another single-level interbody fusion. In the cases of multiple level pathology, we preferred to perform a single level of arthroplasty and fusion, instead of multiple levels of arthroplasty due to lack of results of outcome yet. To maximize benefits of arthroplasty, arthroplasties were usually performed at the upper level of arthrodesis. Single level interbody fusions were performed in 10 cases at the just lower adjacent level and in 2 cases at the non-consecutive level. The preoperative symptoms and signs were radiculopathy in 16 patients (34.8%), myelopathy in 4 patients (8.7%), combined radiculopathy and myelopathy in 6 patients (13.0%) and radiculopathy with axial neck pain in 20 patients (43.5%). The duration of preoperative history averaged 17.2 months (range 0.25 to 120 months). All patients underwent an anterior cervical discectomy followed by the implantation of the Bryan prosthesis. We followed the standard surgical technique using parallel drilling of endplates and circular milling to a predetermined size (14–18 mm). Clinical outcome was assessed at each time frame using the visual analogue scale (VAS) and patients’ self-administered neck disability index (NDI).

Surgical indication Inclusion criteria included single or two level disease between C3=4 and C6=7, in patients with either radiculopathy or myelopathy, not responding to conservative treatment. Cases of predominantly anterior compression of the cervical spinal cord or nerve roots are good candidates. Exclusion criteria included infection, metabolic bone disease,

Prior to surgery, patients were evaluated using static and dynamic cervical spine radiographs in addition to computerized tomography (CT) and magnetic resonance (MR) imaging. Static neutral lateral radiographs were used to assess cervical sagittal balance, and anterior-posterior (AP) radiographs were used to exclude abnormal coronal alignment. Dynamic cervical spine radiographs were obtained to confirm movement at the level of the proposed surgery and at adjacent levels. The presence of movement at the level for which the procedure was proposed was prerequisite for arthroplasty. Computerized tomography was used to assess the size of the disc prosthesis and also aided in the evaluation of facet disease. The CT was also valuable in visualizing the location of the foramen transversarium, the morphology of the vertebral body, and the location of foraminal stenosis. Preoperative MR imaging was performed on all patients to determine the origin of radiculopathy or myelopathy. Static and dynamic cervical spine radiographs were obtained on 1 day, 1 week, and then at 1, 6, 12, and 18 months postoperatively. The patient population of radiological follow-up obtained at each time frame were as follows: 1 day (46 patients), 1week (46 patients), 1 month (46 patients), 6 month (23 patients), 12 month (10 patients), 18 month (5 patients). The sagittal alignment of the cervical spine was calculated using the Cobb’s angle method between the inferior margin of C2 and C7 vertebral bodies in a neutral position. Lordosis is shown as a negative value and kyphosis is shown as a positive value. The range of movement of the cervical spine was defined as the difference in the Cobb’s angle between the full flexion and extension as shown in lateral radiographs (Fig. 1). To analyze movement at the level of the proposed arthroplasty, we examined the functional spinal unit (FSU) angle [28]. The FSU angle is formed by lines drawn at the superior margin of the superior vertebral body and at the inferior margin of the inferior body. The FSU angle shows lordosis as a negative value, and kyphosis as a positive value. A neutral FSU angle and the range of movement (ROM) of the FSU angle were calculated in the static and dynamic lateral radiographs (Fig. 2). Radiographs were also analyzed to determine the shell angle, defined as the angle of the disc space formed postoperatively by the shells of the prosthesis as the complementary measure of FSU angle.

Fig. 1. The range of movement (ROM ) of the cervical spine was defined as the difference in the Cobb’s angle between the full flexion and extension as shown in lateral radiographs

945

Clinical and radiological results following cervical arthroplasty

Fig. 2. The range of movement (ROM ) of functional segmental unit (FSU) was calculated in the dynamic lateral radiographs using quantitative measurement software In the analysis of the neutral angle and ROM of the adjacent level, we examined the upper level of the proposed arthroplasty using the Cobb’s angle method at the segment. We measured the angles with quantitative measurement analysis software, PACS workstation (Centricity 2.0, General electrics medical systems, Milwaukee, WI, USA), which uses extrapolative algorithms to calculate the intersecting angle between two lines drawn by the investigator. To fairly assess the margins of the vertebral body, all angles were measured three times and a mean angle was determined. The same investigator performed all computerized measurements. Standard statistical analysis was used for this study. To evaluate the differences between the sagittal alignment, ROM, and the clinical outcomes at each time point, we used a parametric paired sample t-test and linear regression analysis using SPSS software for Windows (SPSS Inc., Chicago, IL, USA, Ver. 12.0).

Results

mean preoperative score of 74% to a postoperative score of 24% for those patients who reached the 1 year postoperative follow-up. The VAS scores of radiating pain improved significantly in the late follow-up results from a mean preoperative score of 8.15 to 1.35 at 1 year postoperatively. The VAS scores of neck pain decreased from a mean preoperative score of 6.5 to 3.8 at 1 year, according to the paired sample t-test. As of the last follow up, neck pain had worsened in 4 patients: the VAS score of neck pain increased 5.32 to 6.90, the NDI scores increased from 67% to 75% in those 4 patients (mean follow up period was 13.8 months).

Surgical outcome Radiological analysis Preoperative clinical diagnosis was herniated cervical disc in 39 patients (84.8%, Soft disc 23=Hard disc 16), cervical stenosis in 6 patients (13.0%, Hard disc 4=OPLL 2), and Klippel-Feil syndrome in 1 patient (2.2%). The levels of surgery included the following: C4–5 (4 patients, 8.7%), C5–6 (32 patients, 69.6%), and C6–7 (10 patients, 21.7%). The size of the Bryan discs used varied as follows: 14 mm (3 patients), 15 mm (10 patients), 16 mm (17 patients), 17 mm (11 patients), 18 mm (5 patients). The mean duration of surgery (excluding time for induction of anesthesia) was 2.3 hours (range 1.4–3.5 hours). Mean blood loss during the operation was 208 ml.

ROM of the whole cervical spine The ROM of the cervical spine was calculated using the difference of C2–7 Cobb’s angles at full flexion and full extension in lateral radiographs. The mean preoperative ROM was 54.0
14.7degrees (mean
standard deviation). The ROM was smaller than the preoperative value in early follow up evaluation (1 month, 36.5
11.0 degrees), but recovered to that of the preoperative value in later follow up (1 year, 55.1
18.5 degrees). The differences in ROM were not statistically significant (Fig. 3). ROM of treated segment

Clinical outcome The NDI scores decreased significantly in late followup evaluations (paired sample t-test, p< 0.05) from a

The ROM of the treated segment was assessed at both the FSU and the Shell angle. All patients showed movement at the treated segment after the procedure. The

946

Fig. 3. The range of movement (ROM ) of the whole cervical spine (C2–7) and FSU were smaller than preoperative values in early follow up evaluation but were restored to preoperative values with one year of the procedure. There were no statistical differences in ROM

mean preoperative ROM of the FSU was 13.3
5.7 degrees. As shown in the ROM change of the whole cervical spine, the ROM of the FSU was more restricted in the early follow up evaluation (1 month, 9.3
3.7 degrees) but reverted back to the preoperative value at a later follow up (1 year, 14.4
4.5 degrees). The treated segment ultimately showed preservation of movement when compared to preoperative levels (Fig. 3). The ROM of the shell angle at the treated level also showed postoperative preservation and mimicked the pattern of change of the FSU (Fig. 4).

D. H. Yoon et al.

Fig. 5. The range of movement (ROM) of the adjacent segment did not increase significantly after the procedure

15.7
4.3 degrees at 1 year postoperatively (Fig. 5). There is no statistical difference in the ROM of adjacent segments after the procedure. Sagittal alignment of cervical spine

The ROMs of the upper segments were 13.1
3.9 degrees preoperatively, 9.0
3.2 degrees at 1 month and

The mean preoperative cervical curvature was lordotic at 14.8
11.2 degrees. In the early postoperative period, cervical curvature showed evidence of kyphotic changes with a measured angle of 7.4
10.8 degrees at 1 week, and 10.9
7.9 degrees at 1 month. At later follow-up, cervical curvature recovered lordosis with angles of 12.6
6.3 degrees at 1 year. There was ultimately no statistical difference in sagittal alignment. Sagittal alignment of the FSU was maintained during the follow-up period (Fig. 6). There was a significant positive correlation between preoperative and postopera-

Fig. 4. The range of movement (ROM) of shell angle showed postoperative movement preservation in the treated level and the consecutive change pattern of FSU

Fig. 6. Sagittal alignment of cervical spine. In the early postoperative period, cervical curvature became kyphotic, but in the later period, cervical curvature recovered lordosis at 1 year postoperatively

ROM of adjacent segment

Clinical and radiological results following cervical arthroplasty

Fig. 7. There was a significant positive correlation between preoperative and postoperative cervical curvature, as shown with a linear correlation study of the early period after the procedure. (Y ¼ 2.970 þ 0.686X, r2 ¼ 0.498, p < 0.05)

tive cervical curvature, as shown by a linear correlation study of the early period after the procedure (Fig. 7). Complications We experienced one case of an acute subdural haematoma, ventral to the spinal cord at the C1–T4 levels. This caused transient quadruplegia in the immediate postoperative period. The patient’s neurological deficits resolved completely within 2 hours of the procedure with conservative management including high-dose steroid medication. Other common complications of anterior cervical arthrodesis, such as device migration, infection, neurological damage, or fusion, did not occur in our study. Discussion The ultimate aim of cervical arthroplasty is to avoid degeneration of adjacent segments by maintaining the range of movement [2]. Patients who have undergone anterior cervical fusion may show radiographic and clinical evidence of progressive degeneration of adjacent segments. Long-term radiographic follow-up of patients with anterior cervical fusion has demonstrated hypermobility and degenerative changes in the non-fused segments of the spine including disc space narrowing, endplate sclerosis, and osteophyte formation [3, 12, 20, 31, 32]. Rates of degeneration of segments adjacent to the fusion level range from 0 to 81%, as shown in previous radiological studies [18, 19]. Although several authors have demonstrated radiographic degeneration

947

with no clinical correlation [8, 13, 23, 27], Gore and Sepic [20] found an association between cervical pain and the progression of spondylosis. Hilibrand et al. [21] performed a retrospective analysis of 409 anterior cervical arthrodeses performed in 374 patients and found that symptomatic adjacent segment disease occurred at a relatively constant rate of 2.9% per year, with a projected survival rate of 25.6% at 10 years. The reported rates of symptomatic adjacent segment disease vary widely, from 2.2% to 14% with varying follow up durations [3, 6, 22]. This is the important background of emergence of cervical arthroplasty as an alternatives to arthrodesis that offers the promise of restoring normal spinal movement and reduce a kinematic strain on adjacent segments. To maintain cervical movement after discectomy, several devices have been employed, including balland-socket joint [11, 31], Prestige [29, 30], the Bryan disc, and Prodisc C [4, 5]. Favorable results have been reported about procedures using the Bryan Disc in Europe [14, 17, 28]. Early clinical results from European trials of single- or double-level prosthesis implantation demonstrated that cervical movement was preserved in 88% of patients treated with singlelevel and in 86% of those with double-level prostheses at 1 year. Clinical outcomes were rated as excellent, good, or fair in more than 90% of patients at 1 year postoperatively [17]. In our series, young patients who have a high lifetime risk of adjacent segment disease are good candidates for prosthesis. In older patients, arthroplasty may be better suited for those with an increased risk of adjacent segment disease (i.e. spondylotic changes are present at adjacent or other spinal levels). The artificial disc may also be indicated for patients who have previously undergone an arthrodesis and present with a degenerated disc at an adjacent or distant level. Preoperative evidence of segmental movement is an important prerequisite for arthroplasty. If movement is limited, if there is evidence of advanced facet disease, or if neck pain is the predominant symptom, arthroplasty is excluded due to limited benefit. But, in our series, only one arthroplasty was performed on a nonmobile segment. The patient with Klippel-Feil syndrome without segmental movement underwent arthroplasty to restore it and prevent degeneration of the remaining upper functional unit. In our relatively short-term series, 100% of patients with a single-level arthroplasty demonstrated preservation of cervical mobility at the treated level and its

948

maintenance in adjacent segments. ROM of the treated segment, the whole cervical spine and of adjacent segments was restricted in the early postoperative period due to neck pain or patients’ non-compliance to movements as an early response to surgery. ROM was restored to preoperative levels when evaluated over longer periods and there are no statistical differences. The maintenance in the ROM of adjacent segments in the late period indirectly reflect the possibility of preventing adjacent segment disease after arthroplasty. Although recent reports and our results show successful preservation of movement in the treated segment and the prevention of hypermobility of adjacent segments, we must emphasize the importance of postoperative sagittal alignment of the whole cervical spine and its relation to long-term clinical outcome, especially for axial neck pain and adjacent segment disease. It is controversial whether cervical kyphosis influences clinical outcome after anterior cervical discectomy with or without fusion [1, 25]. Although the ROM is an important feature of an artificial disc, it is only a single measure of spinal biomechanics. The effect of the artificial disc on angulation at the treated level and on the overall spinal alignment may be important for longterm clinical outcome and the rate of adjacent segment disease. In a retrospective study of 42 patients after anterior cervical discectomy and fusion, Katsuura et al. [25], found that degeneration of adjacent levels was significantly associated with loss of physiological cervical lordosis. They speculated that fusion in kyphosis may load posterior slipping forces on adjacent vertebral levels and may cause higher loads on the posterior column than a fusion in lordosis. The implication is that physiological lordotic fixation is required to prevent neurological deterioration. In our series, there is a statistically significant correlation between preoperative and postoperative cervical alignment. Overall early postoperative cervical alignment became more kyphotic than segmental alignment compared to preoperative values. The C2–7 neutral Cobb’s angle became more kyphotic in the early postoperative period and nearly returned to the preoperative values within a year. Patients experienced a loss of lordosis of about 3.8 degrees at 1 month postoperatively, and 2.1 degrees at 1 year postoperatively. The alignment of the FSU was preserved for 1 year following the procedure. This result was contrary to the results of Pickett et al. who found that the FSU angle became more kyphotic than the C2–7 neutral Cobb’s

D. H. Yoon et al.

angle [28]. In our study, 4 patients suffered aggravated axial neck pain at the last follow up, but precise evaluation for correlation between postoperative alignment and clinical outcome is under investigation (unpublished data). We hypothesized that the early kyphotic changes can be the result of neck pain and may be an early response to surgery. Late kyphotic changes may be the result of various factors, as described below. 1) Over-milling at the dorsal endplate. In the milling step of arthroplasty, there is the possibility of overmilling at the dorsal portion of endplate because the pivot of lever action during the endplate milling is located near and outside of the center of the milling. The anterior portion of the Bryan disc can be affected by more compressive forces than the posterior portion as experienced with normal posture of the cervical spine. This may result in the decrease in the anterior disc space height. 2) The angle of Bryan disc insertion. As mentioned in the literature [28], the angle of disc insertion can significantly alter the orientation of the prosthesis. This is related to bone removal and endplate preparation for the prosthesis. The product monograph of the Bryan cervical disc system suggests that the disc space angle be calculated by measuring the angle between the plumb line and a line connecting the posterior inferior corner of the caudal vertebral body and the posterior superior corner of the cephalad vertebral body. The results obtained using this method can vary depending on the level of the target disc space and the method of intra-operative positioning. Using another method to calculate the angle of insertion, the prosthesis is inserted parallel to the angle of the disc space. Changing the angle of disc insertion, however, is only effective in avoiding kyphosis in the prosthesis endplates. We prefer the second method for preventing kyphotic changes. Further studies will be needed to address the appropriate method for calculating the angle of disc insertion. 3) Structural absence of lordosis in the Bryan disc. Because of the passive nature of the nucleus and uniformity in the pressure resistance around the prosthesis, intradisc pressure is mainly affected by cervical weight balance and the axis of segmental movement. The absence of support of the anterior portion of the Bryan disc can cause a loss of lordosis and may allow hyperflexion to occur. If the Bryan disc has a another anterior compartment having more resistance to the compressive force rather than pos-

949

Clinical and radiological results following cervical arthroplasty

terior compartment or is designed with a feature of lordosis (manufactured with various graded lordotic angle), maintenance or correction of the cervical lordotic angle could be achieved. 4) Surgical procedure to remove the posterior longitudinal ligament (PLL). Destruction of anterior column of cervical spine and destruction of restrictive force to hyperflexion of cervical spine by complete removal of PLL during operation may play a role in increasing the kyphotic change. Most surgeons would advocate removal of the PLL to assure adequate decompression. But we prefer to preserve the PLL as much as possible after partial removal of PLL enough for adequate decompression in order to maintain the force that restricts hyperflexion. 5) Pre-existing kyphosis. In patients with preoperative focal kyphosis, there is a risk that the kyphosis may worsen after insertion of the disc. The disc preserves the existing mobility of the spine, which has been altered by the underlying degenerative process. It thus maintains an abnormal condition rather than correcting it. Careful patient selection and attention to the angle of insertion should reduce the risk of postoperative kyphotic deformity [28]. Our results indicate that cervical arthroplasty provides favorable clinical and radiological outcomes for at least a relatively short follow up period. Our clinical results are similar to those described in previous studies of the same procedure. The potential benefit of arthroplasty over anterior cervical fusion will be determined with the aid of late follow-up evaluations, in which the rates of adjacent segment disease and repeated surgery are examined. Prospective study to compare the efficacy and outcome of several devices of arthroplasty and to examine the superiority over fusion will be needed in the near future. Conclusion Arthroplasty using the Bryan disc appears to be safe and provided a favorable clinical and radiological outcome in our study. Long-term follow up studies are required to prove its efficacy and its ability to prevent adjacent segment disease. Postoperative kyphosis following arthroplasty could be prevented by elucidating the biomechanical properties of the Bryan disc. Further studies are needed to determine whether postoperative kyphosis following arthroplasty is associated with a poor clinical outcome or with adjacent segment disease.

References 1. Abd-Alrahman N, Dokmak AS, Abou-Madawi A (1999) Anterior cervical discectomy (ACD) versus anterior cervical fusion (ACF), clinical and radiological outcome study. Acta Neurochir (Wien) 141: 1089–1092 2. Anderson PA, Rouleau JP (2004) Intervertebral disc arthroplasty. Spine 29: 2779–2786 3. Baba H, Furusawa N, Imura S, Kawahara N, Tsuchiya H, Tomita K (1993) Late radiographic findings after anterior cervical fusion for spondylotic myeloradiculopathy. Spine 18: 2167–2173 4. Bertagnoli R, Duggal N, Pickett GE, Wigfield CC, Gill SS, Karg A, Voigt S (2005) Cervical total disc replacement, part two: clinical results. Orthop Clin North Am 36: 355–362 5. Bertagnoli R, Yue JJ, Pfeiffer F, Fenk-Mayer A, Lawrence JP, Kershaw T, Nanieva R (2005) Early results after ProDisc-C cervical disc replacement. J Neurosurg Spine 2: 403–410 6. Bohlman HH, Emery SE, Goodfellow DB, Jones PK (1993) Robinson anterior cervical discectomy and arthrodesis for cervical radiculopathy. Long-term follow-up of one hundred and twenty-two patients. J Bone Joint Surg Am 75: 1298–1307 7. Bose B (1998) Anterior cervical fusion using Caspar plating: analysis of results and review of the literature. Surg Neurol 49: 25–31 8. Cherubino P, Benazzo F, Borromeo U, Perle S (1990) Degenerative arthritis of the adjacent spinal joints following anterior cervical spinal fusion: clinicoradiologic and statistical correlations. Ital J Orthop Traumatol 16: 533–543 9. Cinotti G, David T, Postacchini F (1996) Results of disc prosthesis after a minimum follow-up period of 2 years. Spine 21: 995–1000 10. Clements DH, O’Leary PF (1990) Anterior cervical discectomy and fusion. Spine 15: 1023–1025 11. Cummins BH, Robertson JT, Gill SS (1998) Surgical experience with an implanted artificial cervical joint. J Neurosurg 88: 943–948 12. DePalma AF, Rothman RH, Lewinnek GE, Canale ST (1972) Anterior interbody fusion for severe cervical disc degeneration. Surg Gynecol Obstet 134: 755–758 13. Dohler JR, Kahn MR, Hughes SP (1985) Instability of the cervical spine after anterior interbody fusion. A study on its incidence and clinical significance in 21 patients. Arch Orthop Trauma Surg 104: 247–250 14. Duggal N, Pickett GE, Mitsis DK, Keller JL (2004) Early clinical and biomechanical results following cervical arthroplasty. Neurosurg Focus 17: E9 15. Enker P, Steffee A, McMillin C, Keppler L, Biscup R, Miller S (1993) Artificial disc replacement. Preliminary report with a 3-year minimum follow-up. Spine 18: 1061–1070 16. Goffin J, Casey A, Kehr P, Liebig K, Lind B, Logroscino C, Pointillart V, Van Calenbergh F, van Loon J (2002) Preliminary clinical experience with the Bryan Cervical Disc Prosthesis. Neurosurgery 51: 840–845; discussion 845–847 17. Goffin J, Van Calenbergh F, van Loon J, Casey A, Kehr P, Liebig K, Lind B, Logroscino C, Sgrambiglia R, Pointillart V (2003) Intermediate follow-up after treatment of degenerative disc disease with the Bryan Cervical Disc Prosthesis: single-level and bi-level. Spine 28: 2673–2678 18. Goffin J, van Loon J, Van Calenbergh F, Plets C (1995) Long-term results after anterior cervical fusion and osteosynthetic stabilization for fractures and=or dislocations of the cervical spine. J Spinal Disord 8: 500–508; discussion 499 19. Gore DR, Gardner GM, Sepic SB, Murray MP (1986) Roentgenographic findings following anterior cervical fusion. Skeletal Radiol 15: 556–559 20. Gore DR, Sepic SB (1984) Anterior cervical fusion for degenerated or protruded discs. A review of one hundred forty-six patients. Spine 9: 667–671

950

D. H. Yoon et al.: Clinical and radiological results following cervical arthroplasty

21. Hilibrand AS, Carlson GD, Palumbo MA, Jones PK, Bohlman HH (1999) Radiculopathy and myelopathy at segments adjacent to the site of a previous anterior cervical arthrodesis. J Bone Joint Surg Am 81: 519–528 22. Hilibrand AS, Yoo JU, Carlson GD, Bohlman HH (1997) The success of anterior cervical arthrodesis adjacent to a previous fusion. Spine 22: 1574–1579 23. Hunter LY, Braunstein EM, Bailey RW (1980) Radiographic changes following anterior cervical fusion. Spine 5: 399–401 24. Jacobs B, Krueger EG, Leivy DM (1970) Cervical spondylosis with radiculopathy. Results of anterior diskectomy and interbody fusion. Jama 211: 2135–2139 25. Katsuura A, Hukuda S, Saruhashi Y, Mori K (2001) Kyphotic malalignment after anterior cervical fusion is one of the factors promoting the degenerative process in adjacent intervertebral levels. Eur Spine J 10: 320–324 26. Matsunaga S, Kabayama S, Yamamoto T, Yone K, Sakou T, Nakanishi K (1999) Strain on intervertebral discs after anterior cervical decompression and fusion. Spine 24: 670–675 27. McGrory BJ, Klassen RA (1994) Arthrodesis of the cervical spine for fractures and dislocations in children and adolescents. A longterm follow-up study. J Bone Joint Surg Am 76: 1606–1616 28. Pickett GE, Mitsis DK, Sekhon LH, Sears WR, Duggal N (2004) Effects of a cervical disc prosthesis on segmental and cervical spine alignment. Neurosurg Focus 17: E5 29. Robertson JT, Metcalf NH (2004) Long-term outcome after implantation of the Prestige I disc in an end-stage indication: 4-year results from a pilot study. Neurosurg Focus 17: E10 30. Traynelis VC (2004) The Prestige cervical disc replacement. Spine J 4: 310S–314S 31. Wigfield C, Gill S, Nelson R, Langdon I, Metcalf N, Robertson J (2002) Influence of an artificial cervical joint compared with fusion on adjacent-level motion in the treatment of degenerative cervical disc disease. J Neurosurg 96: 17–21 32. Wu W, Thuomas KA, Hedlund R, Leszniewski W, Vavruch L (1996) Degenerative changes following anterior cervical discectomy and fusion evaluated by fast spin-echo MR imaging. Acta Radiol 37: 614–617

Comments Anthroplasty of the cervical spine still remains a new technology about which we do not know yet where it is going to bring us and our patients on the long run. We are still in a phase in which we need to collect as much clinical data as possible, provided that this data derives from well designed clinical studies. The manuscript by Yoon et al. is an important study, although the clinical and radiological results have to be interpreted with caution since the follow-up is very short. It is surprising to me that the authors consider bi-level arthroplasty as ‘‘too experimental’’ to be applied in clinical practice: there is nowadays enough information in the literature for concluding that bi-level disc replacement is feasible and safe.

Personally I have some concerns about treating spondylotic myelopathy or myelopathy due to OPLL with artificial disc replacement. Indeed, the concept of arthroplasty in the cervical spine has to do with prevention of acceleration of adjacent level degeneration, as it is often seen after discectomy and interbody fusion. However, myelopathy is a disease of a different magnitude than radiculopathy: in myelopathy the disease should be definitely interrupted at the operated level and adjacent level issues are of secondary concern. There are some indications that in case an arthroplasty is performed in spondylotic myelopathy or OPLL posterior osteophytes or ossification of the posterior ligament at the operated level might develop again, due to the ongoing motion. In case of a narrow spinal canal, this might lead to recurrent signs and symptoms of myelopathy. The situation is different for radiculopathy. Yoon et al. performed a very nice study on ROM and alignment of the global cervical spine and of the operated level. It is interesting to notice how they were able to avoid kyphosis at the target level, as it is sometimes seen after insertion of a Bryan disc (reference 28). I am also convinced that kyphosis can be avoided by slightly modifying the original surgical technique: since the time I started inserting the sagittal wedge and the milling guide parallel to the cranial endplate of the caudal vertebral body and did not compress the disc anymore too firmly when the fluid is injected, I did not encounter any kyphosis with the Bryan disc. Patients with pre-existing important kyphosis should of course not be treated with an artificial disc. It is also interesting and satisfying to notice that the ROM of the operated level returned to its normal value after some initial postoperative limitation of motion. The suggestion of the authors that preservation of the posterior ligament might maintain the force that limits hyperflexion is interesting: however, one has to keep in mind that the practice of leaving the posterior ligament intact might lead to insufficient nerve root compression. Personally I therefore always remove the posterior longitudinal ligament. Jan Goffin Leuven Yoon et al. provide a retrospective clinical and radiological outcome study on 46 patients undergoing Bryan disc cervical arthroplasty with a mean follow-up of 48 months. The document good clinical outcomes with preserved motion in but showed a worrying trend towards to kyphosis that corrected after 12 months. This is an interesting study as it documents the radiological course of sagittal imbalance and the authors have reviewed potential causes, many of which may be devicespecific. As a plethora of arthroplasty devices hit the world market similar careful and thoughtful analysis of both clinical and radiological outcomes should be sort as unforeseen problems such as kyphosis are identified that may not be immediately identifiable. Lali Sekhon Reno, Nevada, U.S.A.

Correspondence: Seong Yi, Department of Neurosurgery, Yonsei University, College of Medicine, Seoul 120-752, Korea. e-mail: [email protected]