`
`2003;66:676-681
`
`Shih-Tien Wang1
`Vijay K. Goel2
`Shinichiro Kubo3
`Woosung Choi4
`Justin K. Coppes5
`Chien-Lin Liu6
`Tain-Hsiung Chen1
`
`1 Department of Orthopedics and
`Traumatology, Taipei Veterans General
`Hospital and National Yang Ming
`University, Taipei, Taiwan, R.O.C.
`2 Spine Research Center, University of
`Toledo and Medical College of Ohio,
`Toledo, OH, U.S.A.;
`3 Department of Orthopedic Surgery,
`Miyazaki Medical College, Miyazaki,
`Japan;
`4 Department of Orthopaedics, Tae Jeon St.
`Mary’s Hospital, The Catholic University
`of Korea, Seoul, Korea; and
`5 Department of Biomedical Engineering,
`University of Iowa, Iowa City, IA, U.S.A.
`6 Department of Surgery, School of
`Medicine, National Yang Ming
`University, Taipei, Taiwan, R.O.C.
`
`Key Words
`Bagby and Kuslich (BAK) cage;
`
`biomechanics;
`
`cadaver model;
`
`stability;
`
`supplementary posterior instrumenta-
`
`tion
`
`Original Article
`
`Comparison of Stabilities between
`Obliquely and Conventionally Inserted
`Bagby and Kuslich Cages as Posterior
`Lumbar Interbody Fusion in a Cadaver
`Model
`
`Background. The Bagby and Kuslich (BAK) cage as posterior lumbar interbody fu-
`
`sion (PLIF) is reported to give satisfactory results in restoring spinal stability. More-
`
`over, correction by obliquely inserting a single BAK cage has the advantages of re-
`
`ducing exposure, precise implantation, and lower cost. However, biomechanical data
`
`on this procedure are not abundant. This study was designed to compare the stability
`
`imparted by the cages placed using an oblique and posterior approaches and to deter-
`
`mine the effects of supplementary posterior instrumentation.
`
`Methods. After affixing nine human cadaveric spines (L2-S1) within a testing frame,
`
`load testing in several clinically relevant modes was performed sequentially for the
`
`intact and the following procedures across the L4-5 segments: posterior destabi-
`
`lization, stabilization using 2 parallel BAK cages (CBAK group) or 1 oblique BAK
`
`cage (OBAK group), and additional stabilization with posterior instrumentation.
`
`Spatial locations of vertebral bodies were recorded after each loading step using a
`
`3-D motion measurement system.
`
`Results. Except the OBAK group that had a lower stability in left axial rotation, there
`
`were no significant differences in the stability between both groups in all loading
`
`modes for the stabilization using cages alone. Compared with the intact cases, CBAK
`
`cages provide significant improvement in the stability in 5 displacement modes and
`
`OBAK cage may restore the stabilities of the specimens to the intact state in 5 modes
`
`and provide significant improvement in flexion. Addition of supplementary posterior
`
`instrumentation significantly reduced the angular displacements in both groups.
`
`Conclusions. Both methods of cage insertion have similar stability. Both implanta-
`
`tions, alone or with posterior instrumentation, may improve the stability of the spine,
`
`although posterior instrumentation may further strengthen the stability. The oblique
`
`insertion is more favorable since it requires less exposure, enables precise implanta-
`
`tion, and is less expensive.
`
`The Bagby and Kuslich (BAK) method of lumbar
`
`interbody fusion is a safe and effective technique to
`restore spinal stability through the anterior or posterior
`approach.1,2 In a 2-year follow-up prospective, multi-
`center study, the BAK cage (Sulzer Spine-Tech, Minne-
`
`apolis, Minnesota) as posterior lumbar interbody fusion
`(PLIF) was evaluated to have an overall fusion rate of
`86% with no device-related death and complications in
`12 months after surgery.3 Conventionally, 2 BAK cages
`are inserted from the posterior approach as posterior
`
`Received: November 18, 2002.
`Accepted: July 31, 2003.
`
`Correspondence to: Shih-Tien Wang, MD, Department of Orthopedics and Traumatology, Taipei
`Veterans General Hospital, Taipei 112, Taiwan.
`Fax: +886-2-2874-5839; E-mail: stwang@vghtpe.gov.tw
`
`676
`
`
`
`November 2003
`
`Obliquely vs. Conventionally Inserted BAK Cages
`
`lumbar interbody fusion. Recently, implanting a single
`BAK cage obliquely from a posterior approach to pro-
`vide anterior column support has also been employed.
`This implantation has the advantages of reducing expo-
`sure, precise implantation and lower cost.
`The biomechanical properties of BAK (bilateral ap-
`proach) and SynCage (central approach) have been com-
`pared to find no significant difference in the stabilization
`provided by these 2 designs.4 Moreover, PLIF with a sin-
`gle posterolateral long threaded cage with unilateral
`facetectomy shows to be capable of providing sufficient
`decompression and maintaining most of the posterior el-
`ements in bovine lumbar functional spinal units. In com-
`bination with a facet joint screw, adequate postoperative
`stability was achieved.5 In this study, we employed a ca-
`daver model to compare the stability of the oblique inser-
`tion of a single BAK cage and the conventional insertion
`of two BAK cages in parallel for PLIF across the L4-L5
`segments. In addition, the effects of supplementary pos-
`terior instrumentation were also investigated.
`
`METHODS
`
`Specimen preparation
`Nine intact fresh human cadaver spines (L2-S1)
`were prepared and randomly divided into 2 groups: 4 for
`the conventional insertion of 2 BAK cages (CBAK
`group) and 5 for the oblique insertion of a single BAK
`cage (OBAK group). The bone mineral density of these
`specimens was determined by DEXA (dual energy x-ray
`absorptiometry) scanning to exclude highly degenerated
`and osteoporotic specimens. The soft tissues on each
`specimen were stripped off and the ligamentous struc-
`tures were left intact. Metallic screws were then inserted
`into the vertebral bodies to ensure a secure fixation be-
`tween the vertebral bodies before affixing the superior
`half of the proximal vertebral body and inferior half of
`the distal body before pouring polyester resin. The meth-
`odology of preparing the specimens and testing is
`well-established.6-10
`
`formed according to the protocol in our previous study.6-10
`Each specimen was sequentially tested in the following
`states: (1) intact; (2) destabilization unilaterally on the
`right (hemilaminectomy) by total facetectomy and par-
`tial discectomy across L4-L5 in the OBAK group or
`destabilization by total bilateral laminectomy and dis-
`cectomy at the same level in the CBAK group; (3) stabi-
`lization using an obliquely inserted BAK cage in the
`OBAK group or 2 parallel BAK cages in the CBAK
`group; and (4) additional stabilization using variable
`screw plates (VSP) system (DePuy-AcroMed, Raynham,
`Massachusetts) across the L4-L5 segments in both
`groups. All implements were inserted according to the
`instructions of the manufacturer.
`
`Testing steps
`The three-dimensional load-displacement behavior
`of each of the vertebra was quantified using the Selspot
`II® Motion Measurement System (Innovision Systems,
`Inc., Warren, MI). Loads, in form of pure moments to L2,
`were applied to the spine in 6 degrees of freedom:
`flexion-extension (± 6 Nm), right and left lateral bending
`(± 6 Nm), and right and left torsional loading (± 6 Nm).
`The maximum load was achieved in 5 equal steps. Spa-
`tial location of the specimen was recorded after each
`loading step. To prevent dehydration during preparation
`and testing, specimens was sprayed with 0.9% NaCl so-
`lution.
`
`Statistical analysis
`Since there were only 4 specimens in the CBAK
`group and 5 in the OBAK group, non-parametric tests
`were employed to analyze changes in the angular motion
`for each loading mode. The raw data and the data nor-
`malized with respect to the intact state were analyzed.
`The Kruskal-Wallis test was used to compare the effect
`of cage design and the Friedman test was used to evalu-
`ate the changes in each state of the 2 groups. The critical
`level of significance was 0.05.
`
`Testing procedures
`Mechanical testing on the spine specimens was per-
`
`The mean angular displacements for all 6 load types
`
`RESULTS
`
`677
`
`
`
`Shih-Tien Wang et al.
`
`Journal of the Chinese Medical Association Vol. 66, No. 11
`
`evaluated are summarized in Tables 1. After stabiliza-
`tion, a much larger left axial rotation was found in the
`OBAK group than in the CBAK group (OBAK 1.77° vs.
`CBAK 0.30° p < 0.05). However, no significant differ-
`
`ences were found in the remaining directions (p > 0.05).
`Analyses using the normalized data also showed the
`same patterns in the differences of angular changes be-
`tween the implementation designs (Figs. 1-3).
`
`Table 1. Summary of flexion/extension, lateral bending, and axial rotation motions for the intact and stabilized specimens
`with BAK cages inserted obliquely (OBAK) or conventionally (CBAK) at the L4-L5 lumbar levels of human
`cadaveric specimens
`
`Step
`
`I
`D
`C
`C + I
`
`I
`D
`C
`C + I
`
`I
`D
`C
`C + I
`
`CBAK
`
`OBAK
`
`p
`
`CBAK
`
`OBAK
`
`p
`
`Extension (°)
`2.34 ± 0.78
`2.95 ± 0.96
`1.26 ± 0.78
`0.87 ± 0.99
`Left lateral bending (°)
`2.91 ± 0.88
`3.46 ± 1.04
`0.84 ± 0.36
`0.43 ± 0.24
`Left axial rotation (°)
`1.85 ± 0.71
`2.01 ± 1.41
`0.30 ± 0.22
`0.25 ± 0.07
`
`1.75 ± 0.61
`2.51 ± 0.43
`1.98 ± 1.92
`0.33 ± 0.15
`
`3.18 ± 2.14
`4.22 ± 1.87
`2.75 ± 1.64
`0.74 ± 0.43
`
`1.11 ± 0.64
`1.82 ± 0.17
`1.77 ± 0.72
`0.51 ± 0.33
`
`0.221
`0.221
`0.624
`0.806
`
`0.806
`0.806
`0.221
`0.327
`
`0.221
`0.221
`0.014
`0.327
`
`Flexion (°)
`-4.42 ± 1.60
`-4.95 ± 2.64
`-5.44 ± 3.63
`-8.29 ± 2.61
`-1.39 ± 1.72
`-1.82 ± 1.11
`-0.65 ± 0.48
`-0.90 ± 0.87
`Right lateral bending (°)
`-3.02 ± 0.91
`-2.63 ± 1.53
`-3.77 ± 1.43
`-2.99 ± 1.68
`-0.71 ± 0.56
`-2.23 ± 1.80
`-0.36 ± 0.14
`-0.68 ± 0.36
`Right axial rotation (°)
`-1.46 ± 1.65
`-2.91 ± 1.32
`-0.66 ± 0.27
`-0.33 ± 0.15
`
`-1.49 ± 0.91
`-1.98 ± 1.19
`-0.78 ± 0.42
`-0.44 ± 0.20
`
`1.000
`0.142
`0.806
`0.086
`
`0.806
`0.462
`0.050
`0.327
`
`0.624
`0.221
`0.086
`0.539
`
`Sample size: OBAK n = 5, CBAK n = 4; corresponding to a 6-Nm load step in four different loading modes (I - intact, D -
`destruction, C - cage only, C+I - cage plus instrumentation).
`
`Extension
`
`Flexion
`
`250
`
`200
`
`150
`
`100
`
`%ofintact
`
`50
`
`0
`
`I
`
`D
`
`C
`
`C+I
`
`I
`
`D
`
`C
`
`C+I
`
`Step
`
`Fig. 1. Normalized angular changes in extension and flexion for the CBAK (G) and OBAK (#) cases. Nomenclature used is: I
`- intact, D - destruction, C - cage only, C+I - cage plus instrumentation. Graphs are for the 6-Nm load step and error bars repre-
`sent standard deviations. There was no significant difference in the angular changes between the 2 groups.
`
`678
`
`
`
`November 2003
`
`Obliquely vs. Conventionally Inserted BAK Cages
`
`Although the values after destabilization became
`higher than at the intact stage in general, significant dif-
`ferences in the angular displacements were only ob-
`
`served in the extension, flexion, and right lateral bending
`modes of the CBAK group and the extension mode of the
`OBAK group (p < 0.05). Except in the right axial rotation
`
`Left lateral bending
`
`Right lateral bending
`
`250
`
`200
`
`150
`
`100
`
`%ofintact
`
`50
`
`0
`
`I
`
`D
`
`C
`
`C+I
`
`I
`
`D
`
`C
`
`C+I
`
`Step
`
`Fig. 2. Normalized angular changes in bending motions for the CBAK (G) and OBAK (#) cases. Nomenclature used is: I - in-
`tact, C - cage only, D - destruction, C+I - cage plus instrumentation. Graphs are for the 6-Nm load step and error bars represent
`standard deviations. There was no significant difference in the angular changes between the 2 groups.
`
`Left axial rotation
`
`Right axial rotation
`
`p = 0.014
`
`350
`
`300
`
`250
`
`200
`
`150
`
`%ofintact
`
`100
`
`50
`
`0
`
`I
`
`D
`
`C
`
`C+I
`
`I
`
`D
`
`C
`
`C+I
`
`Step
`
`Fig. 3. Normalized angular changes in axial rotations for the CBAK (G) and OBAK (#) cases. Nomenclature used is: I - intact,
`C - cage only, D - destruction, C+I - cage plus instrumentation. Graphs are for the 6-Nm load step and error bars represent stan-
`dard detions. Although there was no significant difference in the angular changes between the 2 groups in right axial rotation,
`OBAK had a significantly larger angular change in left axial rotation than CBAK.
`
`679
`
`
`
`Shih-Tien Wang et al.
`
`Journal of the Chinese Medical Association Vol. 66, No. 11
`
`mode (p > 0.05), the mean angular displacements be-
`came significantly lower than the intact cases after im-
`plantation of the CBAK cages (p < 0.05). The implemen-
`tation of CBAK cages provides significant improvement
`in the stability of the specimens in 5 displacement
`modes. In the OBAK group, the mean angular displace-
`ment became significantly lower in the flexion mode
`than in the intact cases after cage implantation (p < 0.05).
`However, there were no significant difference in the re-
`maining 5 displacement modes (p > 0.05). The OBAK
`cage may restore the stability of the specimens to the in-
`tact state in 5 modes and provide significant improve-
`ment in flexion. The same patterns in the angular dis-
`placement differences between the intact state and cage
`implementation after destabilization were observed in
`the two implementation designs using the normalized
`data (Table 1 and Figs. 1-3).
`Except in the left axial rotation mode (p > 0.05), the
`displacements became significantly lower than the intact
`after implementing CBAK cages and adding posterior
`instrumentation (p < 0.05). In the OBAK group, signifi-
`cantly lower displacements were observed in all modes
`(p < 0.05) after adding posterior instrumentation. These
`findings indicate the significant improvement in the sta-
`bility of the specimens in both implementation design
`groups after adding the posterior instrumentation (Table
`1 and Figs. 1-3).
`
`DISCUSSION
`
`The BAK cage has been evaluated to be a superior
`interbody fusion device than other graft materials in vivo
`and in vitro using a calf spine model. This implantation
`with posterior instrumentation is found to have the great-
`est stiffness in flexion/extension and axial rotation while
`bone graft alone gives less initial stiffness than that of the
`intact spine, although the results in axial compression
`seem inconclusive.11 Moreover, this cage is reported to
`have similar biomechanical characteristics as the Threaded
`Interbody Fusion Device or SynCage.12 In an in vivo
`study with a sheep thoracic spine model, BAK with bone
`graft or recombinant human bone morphogenetic pro-
`teins was demonstrated to have the same effects on bio-
`
`680
`
`mechanics and histomorphometry as bone graft alone.13
`In a previous study, the stability of these 2 BAK cage
`implantations has been evaluated in 18 bovine lumbar
`functional spinal units. The PLIF with a single post-
`erolateral long threaded cage with unilateral facetectomy
`not only enables sufficient decompression but also main-
`tains most of the posterior elements. Although the single
`cage implantation is stiffer than the two-cage implanta-
`tion in pure compression, flexion, and left and right
`bending, the differences are not significant.5 Although
`the study of functional spinal units may provide valuable
`information on the mechanical properties, the results
`may be different in many ways from those obtained from
`multi-segmental cadaveric spinal models. Moreover, in-
`formation from functional spinal units may not be ap-
`plied directly to explain the multi-segmental motion
`properties. Biomechanical evaluation using multi-seg-
`mental models should be more appropriate for simulat-
`ing the physiologic movements.14
`This study provided a cadaveric spinal model to
`compare between the conventional insertion of 2 BAK
`cages and the oblique insertion of a single BAK cage
`across the L4-L5 segments via a posterior approach. The
`results indicate that both methods of cage insertion, with
`or without supplementary posterior fixation, provided
`similar stability in all loading modes, except that the lat-
`ter method was found to have a much higher degrees of
`left axial rotation than the former in the horizontal plane,
`because the single BAK was inserted oblique by right to-
`tal facetectomy at the right side. Although CBAK im-
`proved the stability of the spine a lot in 5 displacement
`modes, OBAK may restore the spine to the intact state in
`5 modes and help improve flexion. These findings indi-
`cated the usefulness of OBAK in restoring the stability of
`the spine.
`The biomechanical behaviors of implants with or
`without instrumentation have also been evaluated. In a
`comparative study using calf and human cardaveric
`spines on bone graft and Ray cage, increase in flexion,
`lateral bending stiffness and reduced laxity on flexion,
`extension, and lateral bending were observed in both im-
`plants with supplemental posterior plates fixed by pedicle
`screws across the fusion segment.15 In another bio-
`mechanical study on human cadaver spines, the Stratec,
`
`
`
`November 2003
`
`Obliquely vs. Conventionally Inserted BAK Cages
`
`Ray, and Brantigan cages were determined to achieve the
`greatest stabilization in flexion and lateral bending by the
`addition of posterior transpedicular instrumentation.16
`Although the BAK cage may be used as a standalone de-
`vice, the results of this study indicate that the BAK
`cages, when implanted from a posterior approach, may
`provide higher stability with supplementary posterior in-
`strumentation.
`Based on the results of this study, CBAK and OBAK
`have similar stability in the cadaveric spine model. The
`unilateral approach of OBAK might greatly reduce the
`exposure requirements. It also offers the advantage of the
`biomechanics of a construct consisting of anterior col-
`umn support combined with pedicle screws. The imple-
`mentation of a single cage also significantly diminishes
`the cost. In addition, supplementary posterior instrumen-
`tation may also increase the stability in both types of im-
`plantation.
`
`ACKNOWLEDGEMENTS
`
`This study was supported in part by the Sulzer
`Spine-Tech, Inc. and DePuy, Inc. for providing the pedicle
`screw spinal instrumentation.
`
`REFERENCES
`
`Surg 2000;82A:383-93.
`
`5. Zhao J, Hai Y, Ordway N, Park CK, Yuan HA. Posterior lum-
`
`bar interbody fusion using posterolateral placement of a single
`
`cylindrical threaded cage. Spine 2000;25:425-30.
`
`6. Abumi K, Panjabi M, Duranceu J. Biomechanical evaluation
`
`of spinal fixation devices. Part III. Stability provided by six
`
`spinal fixation devices and interbody bone graft. Spine 1989;
`
`14:1239-55.
`
`7. Goel VK, Voo LM, Weinstein JN, Liu YK, Okuma T, Njus
`
`GO. Response of the ligamentous lumbar spine to cyclic bend-
`
`ing loads. Spine 1988;13:294-300.
`
`8. Goel VK, Lim TH, Gwon J, Chen JY, Winterbottom J, Park J,
`
`et al. Effects of rigidity of an internal fixation device: a com-
`
`prehensive biomechanical
`
`investigation. Spine 1991;16:
`
`S155-61.
`
`9. Hitchon PW, Goel VK, Serhan H, Rogge T, Grosland N,
`
`Sairyo K, Torner J. Biomechanical studies on two anterior
`
`thoracolumbar implants in cadaveric spines. Spine 1999;24:
`
`213-8.
`
`10. Lim T, An H, Hong J, Ahn J, You J, Eck J, McGrady L.
`
`Biomechanical evaluation of anterior and posterior fixations
`
`in an unstable calf spine model. Spine 1997;22:261-6.
`
`11. Brodke DS, Dick JC, Kunz DN, McCabe R, Zdeblick TA. Pos-
`
`terior lumbar interbody fusion. A biomedical comparison, in-
`
`cluding a new threaded cage. Spine 1997;22:26-31.
`
`12. Rapoff AJ, Ghanayem AJ, Zdeblick TA. Biomechanical com-
`
`parison of posterior lumbar interbody fusion cages. Spine
`
`1997;22:2375-9.
`
`13. Cunningham BW, Kanayama M, Parker LM, Weis JC, Sefter
`
`JC, Fedder
`
`IL, Mcfee PC. Osteogenic protein versus
`
`autologous interbody arthrodesis in the sheep thoracic spine.
`
`A comparative endoscopic study using the Bagby and Kuslich
`
`1. Weiner BK, Fraser RD. Spine Update. Lumbar interbody
`
`interbody fusion device. Spine 1999;24:509-18.
`
`cages. Spine 1998;23:634-40.
`
`14. Lee CK, Langrana NA. Lumbosacral spinal fusion. A bio-
`
`2. McCafee PC. Interbody fusion cages in reconstructive opera-
`
`mechanical study. Spine 1984;9:574-81.
`
`tions on the spine. J Bone Joint Surg 1999;81A:859-80.
`
`15. Tencer AF, Hampton D, Eddy S. Biomechanical properties of
`
`3. Kuslich SD, Ulstrom CL, Griffith SL, Ahern JW, Dowdle JD.
`
`threaded inserts for lumbar interbody spinal fusion. Spine
`
`The Bagby and Kuslich method of lumbar interbody fusion.
`
`1995;20:2408-14.
`
`History, technique, and 2-year follow-up results of a United
`
`16. Lund T, Oxland TR, Jost B, Cripton P, Grassmann, Etter C,
`
`States prospective, multicenter trial. Spine 1998;23:1267-79.
`
`Nolte L-P. Interbody cage stabilisation in the lumbar spine.
`
`4. Oxland TR, Hoffer Z, Nydegger T, Rathonyi GC, Nolte L-P. A
`
`Biomechanical evaluation of cage design, posterior instru-
`
`comparative biomechanical investigation of anterior lumbar
`
`mentation and bone density. J Bone Joint Surg 1998;80-B:
`
`interbody cages: central and bilateral approaches. J Bone Joint
`
`351-9.
`
`681