spine volume 30, number 11, pp 1275–1282 ©2005 ... · specimens and 20 mm for porcine specimens....

SPINE Volume 30, Number 11, pp 1275–1282©2005, Lippincott Williams & Wilkins, Inc.

Limitations of the Cervical Porcine Spine in EvaluatingSpinal Implants in Comparison With Human CervicalSpinal SegmentsA Biomechanical In Vitro Comparison of Porcine and Human CervicalSpine Specimens With Different Instrumentation Techniques

Rene Schmidt, MD,* Marcus Richter, MD,* Lutz Claes, PhD,† Wolfhart Puhl, MD,* andHans-Joachim Wilke, PhD†

Study Design. Porcine and human cervical spine spec-imens were in vitro biomechanically compared with dif-ferent instrumentation techniques.

Objectives. To evaluate whether subaxial porcine cer-vical spines are a valid model for implant testing in asingle level corpectomy.

Summary of Background Data. Biomechanical in vitrotests are widely used for implant tests, mainly with hu-man spine specimens. The availability of human cadaversis limited and the properties of the specimen regardingage, bone mineral density, and grade of degenerativechanges is inhomogeneous.

Methods. Six porcine and six human cervical speci-mens were loaded nondestructively with pure moments:1) in an intact state; 2) after a corpectomy of C5 andsubstitution by a cage with integrated force sensor; 3)after additional instrumentation with a posterior screwand rod system with: a) lateral mass and b) pediclescrews; 4) after instrumentation with an anterior plate;and 5) with a circumferential instrumentation. The uncon-strained motion and the axial loads occurring in the cor-pectomy gap were measured, as well as the bone mineraldensity of the specimen before testing.

Results. The range of motion in the intact state, as wellas for the different instrumentations, was comparable forflexion-extension. In lateral bending and axial rotation,marked differences in the intact state as well as for pediclescrew instrumentations occurred.

Conclusions. The subaxial porcine cervical spine is apotential model in flexion-extension because of its bio-

mechanical similarity. For lateral bending and axial rota-tion, the marked differences severly restrict the compara-bility.

Key words: porcine cervical spines, implant testing,corpectomy model. Spine 2005;30:1275–1282

Biomechanical in vitro studies should be performed be-fore new implants are used in clinical practice to provethat they are superior or at least comparable to alreadyestablished implant systems. For such tests, human ca-daveric specimens seem to be preferable; however, somedisadvantages follow the use of a human model, e.g.,restricted availability, wide range of interindividualproperties, mainly the age that contributes to differentbiomechanical properties like range of motion (ROM),1

bone mineral density (BMD), grade of degenerativechanges, or mechanical properties of tissues. This makesa standardized testing difficult and complicates the com-parison of results from different studies. These disadvan-tages of human specimens force a search for alternativemodels. Physical models have therefore been suggested,such as the “missing-vertebra”2 model where the ana-tomic similarity is abandoned at the expense of repro-ducibility and standardization. Normally, they are usedfor pure material testing. Animal models for in vitro ex-perimental use, mainly calf3–7 and porcine,8–18 are usedfrequently for in vitro tests, whereas sheep models aremainly for in vivo use.19–34 These models are a compro-mise between anatomic similarity, homogeneity, and re-producibility. Basic studies about the suitability forcalf,35,36 sheep,37–39 and even baboon40,41 spines exist,whereas they are rare for porcine. Yingling et al42 statedthat cervical porcine spines may be useful for studiesabout human lumbar injury mechanics. Grubb et al15

used a test setup with parallel testing of human and por-cine cervical specimens for anterior plates, with compa-rable results. The purpose of this study was to comparehuman and porcine cervical spines with different instru-mentations to provide data about the suitability of por-cine cervical spines for biomechanical in vitro implanttesting.

Materials and Methods

Six human cadaveric specimens, 2 male and 4 female, averageage 80 years (age range, 66–92 years), consisting of C2 to at

From the *Department of Orthopedics and SCI, and †Department ofOrthopedic Research and Biomechanics, University of Ulm, Ulm, Ger-many.Acknowledgment date: August 14, 2003. First revision date: March 25,2004. Second revision date: July 5, 2004. Acceptance date: July 9,2004.Supported by a grant from the AO Foundation, Switzerland (03-W16).The implants were provided by Ulrich Medizintechnik, Ulm, Germany.The device(s)/drug(s) that is/are the subject of this manuscript is/are notFDA-approved for this indication and is/are not commercially avail-able in the United States.Foundation funds were received in support of this work. Although oneor more of the author(s) has/have received or will receive benefits forpersonal or professional use from a commercial party related directlyor indirectly to the subject of this manuscript, benefits will be directedsolely to a research fund, foundation, educational institution, or othernonprofit organization which the author(s) has/have been associated.Address correspondence and reprint requests to Rene Schmidt, MD,Orthopadische Klinik mit Querschnittgelahmtenzentrum der Universi-tat Ulm, Oberer Eselsberg 45, 89081 Ulm, Germany; E-mail:[email protected]

1275

least T1 or T2 as obtained and 6 porcine spines from C2 to T1,age of all specimens about 6 months, weight between 80 and100 kg, were used. The porcine were a hybrid from DeutscheLandrasse (motherrace) and Pietrain pig (fatherrace). The spec-imens were examined and plain radiographs were taken toexclude soft tissue or bone damage and then stored frozen at�20C in triple-sealed plastic bags. After thawing, the muscletissue was carefully removed and all the ligaments and bonystructures were preserved. To prevent dehydration, the speci-mens were kept moist with saline solution. Handling specimensin the above-described manner does not affect their biome-chanical properties.43,44

BMD was measured in the vertebral body and pedicle of C4and C6 by quantitative computed tomography after calibrationof the CT (XCT 960, Stratec GmbH, Birkenfeld, Germany)with a standardized phantom. A constant ratio of 45% of theinner vertebral or pedicle area was defined as trabecular bone.The BMD was obtained in 1-mm-thick slices and reported inmg/cm3.

The cranial and caudal vertebrae were embedded in polym-ethylmethacrylate (Technovit 3040, Heraeus Kulzer GmbH,Wehrheim, Germany). To obtain a better anchorage of thevertebrae in the polymethylmethacrylate, short screws werepartially driven into the vertebrae before embedding. Screwswere inserted in the vertebrae C4 and C6 to fix the motionanalysis system to the specimen, with regard to the screw tra-jectory for the pedicle screws.

The corpectomy model was chosen because it is widely usedfor spinal implant testing with human spines.9,14,45,46 It offersa high-grade instability and good reproducibility and can beused for many different implant types. The segment C5–C6 isoften involved in trauma or degenerative changes and is con-sidered to have the largest ROM,1 and the corpectomy of C5was therefore selected.

The corpectomy C5 was perfomed using rongeurs and ahigh-speed air drill. The posterior longitudinal ligament waspreserved. The corpectomy was at least 15 mm wide for humanspecimens and 20 mm for porcine specimens. After decompres-sion, the cranial and caudal endplate was prepared to accom-modate the cage with the force sensor.

To achieve a good overview of the biomechanical propertiesof cervical porcine spines, we tested several implants: 1) a pos-terior screw and rod system; a) with lateral mass screws, b)with pedicle screws; 2) anterior plating; and 3) a combinationof anterior plating and a posterior screw and rod system (360°instrumentation). The spinal implants consisted of a posteriorscrew and rod system (Neon occipito-cervical system) with4.0-mm cannulated lateral mass and pedicle screws and a4.5-mm rod as well as an anterior plate (Osmium). All implantswere provided from Ulrich Medizintechnik, Ulm, Germany.

In addition, we monitored the forces in the corporectomygap by a force sensor. The force sensor consisted of a miniatureload cell (Miniatur-Druckkraftsensor Typ 8413, Burster Prazi-sionsmesstechnik, Gernsbach, Germany) capable of measuringaxial compressive forces in a range between 0 and 500 N. Theload cell was mounted in a specially modified cage on the basisof a routinely used cage (ADD, Ulrich Medizintechnik, Ulm,Germany) with a pin in the upper cap of the cage, which fittedinto a suitable groove in the lower cage part (Figure 1). Thisconstrained rotational movements and tipping of the cap,therefore preventing rotational forces. Spikes at the superiorand inferior end prevented the cage from slipping. The cage wasadaptable to the desired graft height by the use of two stainless

steel tubes with different lengths and a continuously adjustablescrew thread at the lower end of the cage. The force sensor wasadjusted to 40 N preload while the specimen was mounted inthe spine tester with unconstrained movement of the specimenin all directions. Thereby each specimen could adopt its ownneutral position.

The specimens were mounted in a previously describedspine tester,47 where the caudal vertebrae were rigidly fixed inthe testing apparatus and the cranial vertebra (C2) was fixed ina cardan joint containing integrated stepper motors that couldintroduce pure moments separatly around three axes. Theother 5 of 6 degrees of freedom were free, enabling the speci-mens to move unconstrained. The segmental motion betweenC4–C6 was measured by a high-resolution, noncontacting ul-trasound motion analysis system (Zebris, Isny, Germany, res-olution 0.06°). Nondestructive loads were applied as pure mo-ments in alternating sequences for right/left lateral bending(�Mx), flexion-extension (�My), and right/left axial rotation(�Mz). The tests were performed with � 2.5 Nm for all direc-tions. To precondition the specimens and to minimize vis-coelastic effects, they were tested with 3 cycles and the data ofthe third cycle were evaluated.48 The ROM of the segment wasdetermined for each direction of loading. ROM was defined asthe angular deformation at maximum load. The obtained val-ues were rounded to one decimal place, to account for theresolution of 0.06° of the motion analysis system. The spinalimplants were tested according to the recommendations for thestandardzation of in vitro stability testing of spinal implants.49

The sample size was small and not distributed normally.Thus, we determined the median, instead of the average forROM and BMD. For the BMD, we also determined the rangebetween the minimal and maximal values in relation to themedian, although the distribution of minimal and maximalvalues was not symmetric, to ease the interpretation.

Results

Generally 10% to 30% higher BMD values occurred forthe porcine specimens with overall narrower distributionof the BMD values (Table 1). The median BMD increasesfrom cranial to caudal and achieves highest values in thepedicles for both species.

For lateral bending, the greatest motion occurred forboth species separately in the intact state, with a notice-

Figure 1. Cage with force sensor and additional steel tubes.Human cervical vertebra (artificial) as scale.

1276 Spine • Volume 30 • Number 11 • 2005

able difference in magnitude for the porcine spines (Fig-ure 2). The porcine spines showed intact a larger spread-ing of the individual minimal and maximal value. Thecage had higher values in the porcine group. Lateral massscrews (LMS), pedicle screws (PDS), and the 360° instru-mentations had comparable values with only slight dis-tinctions. The anterior plate (ANP) had for both specieslower values than the cage and higher values than theother instrumentations, but the spreading of the valuesand the median was higher for the human group.

In flexion-extension (Figure 3), the intact values weresimilar with a median of 20.4° for the human spinescompared with 22.4° for the porcine specimens. Againfollowed by the cage, with comparable results for bothspecies. The LMS, PDS, 360°, and the ANP had highervalues and a larger spreading in the human group. Theporcine specimens showed only small spreading, espe-cially when regarding the range where 50% of the valuesare contained (white box in Figure 3). The values forLMS, PDS, and 360° were similar for the porcine spines,whereas a lower median ROM for the 360° was found inthe human group.

In axial rotation (Figure 4), the intact values weredistinct, with definite smaller values for the porcinespines and consequential larger ROM for the cage com-pared with the intact state in the porcine group. These

two instrumentations were again followed by the ante-rior plate, whereby the median value for the ANP wassmaller in the porcine group. The spreading for the an-terior plate was for the middle 50% (white box in Figure4) comparable, only the maximal and minmal valuesshowed a larger spreading for the human group. Thevalues for the LMS and 360° were comparable, withsmaller spreading for the porcine cadavers. For the PDS,the porcine showed higher values and a larger spreadingthan the human specimens.

A qualitative consistent trend for the median axialloads of the force sensor in the corpectomy gap was ob-served for both species (Figure 5).

Discussion

Although the human spine is the “gold-standard” forimplant testing, a lot of disadvantages have to be dealtwith. Especially, availability and homogeneity of thespecimens should be kept in mind. Generally, in ouropinion only three species are reasonable as substitutesfor implant testing concerning availability and measure-ment: calf, sheep, and porcine.

Regarding availability and costs, the here used por-cine spines are advantageous in comparison with humanspines. They are easily available from local slaughter-houses, show great homogenity for age (usually they areall about 6 months) and weight (about 80–100 kg), andare in a controlled state of health. The differences be-tween the genders is small because these slaughter por-cines are castrated shortly (usually within the first week)after birth. Additionally, the young age probably short-ens the influence of the gender. On the other hand, thereis some vagueness about the behavior of a growing or-ganism, especially for the impact of the growth plate. Inthis cartilaginous plate, which separates the endplate andthe vertebral body (Figures 6, 7), ossification starts in thecenter and not as an anular ring apophysis as in hu-

Table 1. Median BMD (in mg/cm3), Rounded RangeBetween Minimum and Maximum Values in % inRelation to the Median BMD

Median HumanBMD (%)

Median PorcineBMD (%)

Vertebra C4 179.9 � 57 247.9 � 10Pedicle C4 363.4 � 42 467.6 � 21Vertebra C6 194.4 � 30 284.9 � 5Pedicle C6 457.6 � 36 500.4 � 20

Figure 2. Boxplot for lateral bend-ing. Box contains the middle half ofthe scores in the distribution. Thelower hinge represents the 25%quartile, the upper hinge the 75%quartile. The line across the box in-dicates the median. The lines aboveand below the box indicate the max-imal and minimal value, respectively.H, human; P, porcine.

1277Limitations of Cervical Porcine Spine • Schmidt et al

mans.50 In our specimen, the growth plate was still un-fused (Figure 7); and although the instrumentations didnot primarily interfere with the growth plate and nomacroscopic damage occurred, the effect on the biome-chanical properties of the porcine specimens is unpre-dictable.

Concerning the anatomy, a relevant issue is differencein size, which is important for the used implants andscrew lengths. The porcine shows advantages against calfand sheep, as the calf shows generally larger differencesin size, whereas the sheep has the largest differences incomparison with humans for the vertebral body height inthe cervical spine39 (Figure 6). However, differences alsooccur for the porcine. The vertebral body height is larger

for the porcine, which results in a larger corpectomy size(median corpectomy height of 3.2 cm for porcine, 2.5 cmfor human) as well as the endplate width (median cor-pectomy width of 2.1 cm for porcine, 1.6 cm for human,data of this study) and depth and the pedicle width(Wilke et al, preliminary data for porcine spines).42,51,52

For the pedicle height, there is still a lack of consistentinformation, but probably the porcine pedicles are alsolarger than human.

The facet joint orientation is different in both, thesagittal and transverse orientation, with greater anglesfor the porcine spines.42,53 The facet orientation deter-mines the pattern of motion54 and enables the segment toresist to torsional loading.42 Biomechanically, this can be

Figure 3. Boxplot for flexion-extension. Box contains the middlehalf of the scores in the distribution.The lower hinge represents the 25%quartile, the upper hinge the 75%quartile. The line across the box in-dicates the median. The lines aboveand below the box indicate the max-imal and minimal value, respectively.H, human; P, porcine.

Figure 4. Boxplot for axial rota-tion. Box contains the middle half ofthe scores in the distribution. Thelower hinge represents the 25%quartile, the upper hinge the 75%quartile. The line across the box in-dicates the median. The lines aboveand below the box indicate the max-imal and minimal value, respectively.H, human; P, porcine.


seen in our study by the difference of the ROM for theintact state in axial rotation (Figure 4). The porcinespines also lack uncinate processes and uncovertebraljoints, which are thought, among other things, to limitlateral bending.1 This could be one reason for the in-verted biomechanical behavior in lateral bending (Figure2), where the human spines showed a smaller ROM inthe intact state. Functionally, it is reasonable because thequadruped porcine needs to flex laterally to look back-ward, in contrast to the biped human, which has to ro-tate.54 The impressing anterolaterally directed processesof the porcine spines seem to have no significant biome-chanical role.17

Tissue property differences, especially for bone andligaments, were rarely object of studies. Aerssens et al55

Figure 5. Median force sensorcurves, thick line porcine, thin linehuman group. Y-axis, load in N; X-axis, moment in Nm.

Figure 6. Vertebrae C5 from the left: calf, sheep, porcine (genuinebone), and human (artificial).


found significant interspecies differences when studyingseveral parameters of trabecular and cortical bone inseven species, including human and pig. Sikoryn andHukins56 saw similarity between human and pig lumbarligamentum flavum, although they could not excludethat the human ligamentum flavum may be stiffer thanthe pig one. Jiang et al57 found in the thoracic pig andhuman spine the supraspinous and interspinous liga-ments similar, but less thick and tough in pig spines.Because of the lack of information, which still exists inthese questions, it is difficult to exactly name differencesand even more difficult to objectify their influence for thebiomechanical behavior.

The BMD in our study showed a more homogeneousdistribution for the porcine spines, with generally highervalues (Table 1). The more varying BMD and the generalllower values in the vertebra for the human group couldhave led to a lower screw tightening with consequentialhigher ROM values. Especially, the anterior plate is mainlydependent on the bone stock of the vertebra. This couldexplain the consistent higher ROM values and largerspreading of values for the anterior plate in all motion di-rections in the human group. As we did not monitor thescrew torque quantitatively, we can only report that thesubjective screw torque was higher and more constant forthe porcine group. In further studies, it would therefore beof importance to determine the screw torque quantitativelyto demonstrate this observation.

The BMD values for the pedicles in both species weregenerally higher, which could have resulted in a moreconsistent stability of the posterior instrumentations,leading to the smaller discrepancy for the LMS, PDS, and360° instrumentations between the human and porcinegroup. Still, the difference leads to higher values and alarger spreading in the human cadavers. An exceptionwas the pedicle screws in lateral bending and axial rota-tion where lower ROM values were found in the humangroup. The pedicles converged for both species to thefront in a transverse plane, but the pedicle inclination issmaller for the porcine cervical spine in the transversalplane (Figure 8), resembling more the human thoracicspine. Quantitative data for pedicle inclination in human

cervical spines exist52,58 but are still missing for porcinecervical spines. A higher pedicle angle, which meansmore converging of the screws, enhances stability,59,60

thereby leading to a higher rotational stability for thehuman spines.

The force sensor median curves (Figure 5) show aconsistent trend, with a higher homogeneity of the singlecurve for the porcine spines, which is indicated by a moresmooth median curve progression. The kinematic for thedifferent implants is consistent with studies by Foley etal61 and Wang et al,62 although we did not find the ex-cessive loads of more than 200 N as Foley et al61 did. Thesimilarity of the kinematic behavior and thereby the sim-ilarity of load acting on the instrumentations for humanand porcine cervical spines can reinforce the use of cer-vical porcine spines for in vitro tests.

The ROM under the instrumentations showed mostlya smaller spreading for the porcine group. This could bein consequence of the more homogeneous properties ofthe porcine spines, which could only be indirectly mea-sured for the BMD, as data, e.g., for the stiffness of liga-ments or the biomechanical influence of bone composi-tion, are either missing or fragmentary. As spinalinstrumentation evaluation models should show a highconsistency of the specimen to enable reproducible re-sults, this seems to a certain degree even out anatomicdifferences. Interesting is also that the uninstrumented

Figure 7. Porcine cervical spine radiograph, see open growth plates.

Figure 8. Vertebrae C5 from above: calf, sheep, porcine, and human.See the different pedicle inclination for porcine and human (black lines).


porcine spine seems to show a comparable or even largervariability in flexibility when compared with humanspines (Figures 2–4). The disadvantage of missing hu-man variability in a sole animal model2 concerning theflexibility thereby may be reduced and still more homo-geneous results can be obtained.

Some major drawbacks follow the use of porcine cer-vical specimens. Sufficient biomechanical similarity be-tween the two species exists only for flexion-extension. Iflateral bending or axial rotation is evaluated, significantdifferences in the ROM for the intact state and with pedi-cle screw devices occur and complicate or eliminate com-parison of the obtained results. Furthermore, only thesubaxial porcine spine can be used, as the upper cervicalspine seems from the gross anatomic differences not to besuitable. The impact of the growth plate and other ana-tomic and biomechanical differences, which cannot beexactly determined, as for other animal models, are alsoto be kept in mind. Whether these disadvantages arebalanced to a certain degree by the similarity in flexion-extension, easy availability, costs, and homogeneity ofspecimens have to be determined for every specific testprotocol. If the mentioned limitations are considered, theporcine subaxial cervical spine can be an option for im-plant testing. Nevertheless, results obtained in vitroshould be verified by clinical studies.

Key Points

● Porcine and human cervical spine specimensshowed a similar trend in stability for different in-strumentation techniques and similar occurring ax-ial loads in flexion-extension.● For lateral bending and axial rotation, markeddifferences for the intact state and pedicle screwinstrumentations occurred.● Porcine cervical spine specimens have the advan-tage of easy availability, low cost, and homoge-neous properties for age, weight, and bone mineraldensity.● Porcine specimens show predominantly asmaller range of the range of motion values.

References

1. White AA, Panjabi MM. Clinical Biomechanics of the Spine, 2nd ed. Phila-delphia: Lippincott, 1990.

2. Panjabi MM. Cervical spine models for biomechanical research. Spine 1998;23:2684–700.

3. Gurr KR, McAfee PC, Shih C-M. Biomechanical analysis of anterior andposterior instrumentation systems after corpectomy. J Bone Joint Surg Am1988;70:1182–91.

4. Shono Y, McAfee PC, Cunningham BW, et al. A biomechanical analysis ofdecompression and reconstruction methods in the cervical spine: emphasison a carbon-fiber-composite cage. J Bone Joint Surg Am 1993;75:1674–84.

5. Kotani Y, Cunningham BW, Abumi K, et al. Biomechanical analysis of cer-vical stabilization systems: an assessment of transpedicular screw fixation inthe cervical spine. Spine 1994;19:2529–39.

6. Sutterlin C, McAfee PC, Warden KE, et al. A biomechanical evaluation ofcervical spinal stabilization methods in a bovine model: static and cyclicalloading. Spine 1988;13:795–802.

7. Eggli S, Schlapfer F, Angst M, et al. Biomechanical testing of three newly

developed transpedicular multisegmental fixation systems. Eur Spine J 1992;1:109–16.

8. Gilbert SG, Johns PC, Chow DC, et al. Relation of vertebral bone screw axialpullout strength to quantitative computed tomographic trabecular bone min-eral content. J Spinal Disord 1993;6:513–21.

9. Stephens GC, Devito DP, McNamara MJ, et al. Short segment transpedicularCotrel-Dubousset instrumentation: a porcine corpectomy model. J SpinalDisord 1993;6:252–5.

10. Asazuma T, Stokes IA, Moreland MS, et al. Intersegmental spinal flexibilitywith lumbosacral instrumentation: an in vitro biomechanical investigation.Spine 1990;15:1153–8.

11. Allan DG, Russell GG, Moreau MJ, et al. Vertebral end-plate failure inporcine and bovine models of spinal fracture instrumentation. J Orthop Res1990;8:154–6.

12. Nasca RJ, Hollis JM, Lemons JE, et al. Cyclic axial loading of spinal im-plants. Spine 1985;10:792–8.

13. Panjabi MM, Duranceau JS, Oxland TR, et al. Multidirectional instabilitiesof traumatic cervical spine injuries in a porcine model. Spine 1989;14:1111–5.

14. Richman JD, Daniel TE, Anderson DD, et al. Biomechanical evaluation ofcervical spine stabilization methods using a porcine model. Spine 1995;20:2192–7.

15. Grubb MR, Currier BL, Shih J-S, et al. Biomechanical evaluation of anteriorcervical spine stabilization. Spine 1998;8:886–92.

16. Gunning JL, Callaghan JP, McGill SM. Spinal posture and prior loadinghistory modulate compressive strength and type of failure in the spine: abiomechanical study using a porcine cervical spine model. Clin Biomech(Bristol, Avon) 2001;16:471–80.

17. Oxland TR, Panjabi MM, Southern EP, et al. An anatomic basis for spinalinstability: a porcine trauma model. J Orthop Res 1991;9:452–62.

18. Russell G, Lavoie G, Evenson R, et al. A reproducible porcine vertebralfracture for biomechanical testing of spinal fixation devices. Clin Orthop1992;267–72.

19. Nagel DA, Kramers PC, Rahn BA, et al. A paradigm of delayed union andnonunion in the lumbosacral joint: a study of motion and bone grafting ofthe lumbosacral spine in sheep. Spine 1991;16:553–9.

20. Kanayama M, Cunningham BW, Weis JC, et al. The effects of rigid spinalinstrumentation and solid bony fusion on spinal kinematics: a posterolateralspinal arthrodesis model. Spine 1998;23:767–73.

21. Blattert TR, Delling G, Weckbach A. Evaluation of an injectable calciumphosphate cement as an autograft substitute for transpedicular lumbar in-terbody fusion: a controlled, prospective study in the sheep model. Eur SpineJ 2003;12:216–23.

22. Blattert TR, Delling G, Dalal PS, et al. Successful transpedicular lumbarinterbody fusion by means of a composite of osteogenic protein-1 (rhBMP-7)and hydroxyapatite carrier: a comparison with autograft and hydroxyapa-tite in the sheep spine. Spine 2002;27:2697–705.

23. Blattert TR, Delling G, Weckbach A. Pediculoscopic assisted transpedicularspongioplasty for interbody fusion of the lumbar spine: an animal experi-ment study of the sheep model. Unfallchirurg 2002;105:680–7.

24. Moore RJ, Fraser RD, Vernon-Roberts B, et al. The biologic response toparticles from a lumbar disc prosthesis. Spine 2002;27:2088–94.

25. Moore RJ, Crotti TN, Osti OL, et al. Osteoarthrosis of the facet joints resultingfrom anular rim lesions in sheep lumbar discs. Spine 1999;24:519–25.

26. Moore RJ, Latham JM, Vernon-Roberts B, et al. Does plate fixation preventdisc degeneration after a lateral anulus tear? Spine 1994;19:2787–90.

27. Moore RJ, Osti OL, Vernon-Roberts B, et al. Changes in endplate vascular-ity after an outer anulus tear in the sheep. Spine 1992;17:874–8.

28. Moore RJ, Vernon-Roberts B, Osti OL, et al. Remodeling of vertebral boneafter outer anular injury in sheep. Spine 1996;21:936–40.

29. Melrose J, Ghosh P, Taylor TK, et al. A longitudinal study of the matrixchanges induced in the intervertebral disc by surgical damage to the annulusfibrosus. J Orthop Res 1992;10:665–76.

30. Osti OL, Fraser RD, Vernon-Roberts B. Discitis after discography: the roleof prophylactic antibiotics. J Bone Joint Surg Br 1990;72:271–4.

31. Osti OL, Vernon-Roberts B, Fraser RD. 1990 Volvo Award in experimentalstudies. Anulus tears and intervertebral disc degeneration: an experimentalstudy using an animal model. Spine 1990;15:762–7.

32. Kandziora F, Scholz M, Pflugmacher R, et al. Experimental fusion of thesheep cervical spine: II. Effect of growth factors and carrier systems on inter-body fusion. Chirurg 2002;73:1025–38.

33. Kandziora F, Schollmeier G, Scholz M, et al. Influence of cage design on inter-body fusion in a sheep cervical spine model. J Neurosurg 2002;96:321–32.

34. Kandziora F, Pflugmacher R, Schafer J, et al. Biomechanical comparison ofcervical spine interbody fusion cages. Spine 2001;26:1850–7.

35. Cotterill PC, Kostuik JP, D’Angelo G, et al. An anatomical comparison of thehuman and bovine thoracolumbar spine. J Orthop Res 1986;4:298–303.


36. Wilke HJ, Krischak S, Claes L. Biomechanical comparison of calf and humanspines. J Orthop Res 1996;14:500–3.

37. Wilke HJ, Kettler A, Claes LE. Are sheep spines a valid biomechanical modelfor human spines? Spine 1997;22:2365–74.

38. Kandziora F, Pflugmacher R, Scholz M, et al. Comparison between sheepand human cervical spines: an anatomic, radiographic, bone mineral density,and biomechanical study. Spine 2001;26:1028–37.

39. Wilke HJ, Kettler A, Wenger KH, et al. Anatomy of the sheep spine and itscomparison to the human spine. Anat Rec 1997;247:542–55.

40. Tominaga T, Dickman CA, Sonntag VK, et al. Comparative anatomy of thebaboon and the human cervical spine. Spine 1995;20:131–7.

41. Dickman CA, Crawford NR, Tominaga T, et al. Morphology and kinematicsof the baboon upper cervical spine: a model of the atlantoaxial complex.Spine 1994;19:2518–23.

42. Yingling VR, Callaghan JP, McGill SM. The porcine cervical spine as amodel of the human lumbar spine: an anatomical, geometric, and functionalcomparison. J Spinal Disord 1999;12:415–23.

43. Panjabi MM, Krag M, Summers D, et al. Biomechanical time-tolerance offresh caderveric human spine specimens. J Orthop Res 1985;3:292–300.

44. Wilke HJ, Jungkunz B, Wenger K, et al. Spinal segment range of motion as afunction of in vitro test conditions: effects of exposure period, accumulatedcycles, angular- deformation rate, and moisture condition. Anat Rec 1998;251:15–9.

45. Traynelis VC, Donaher PA, Roach RM, et al. Biomechanical comparison ofanterior Caspar plate and three-level posterior fixation techniques in a hu-man cadaveric model. J Neurosurg 1993;79:96–103.

46. Do Koh Y, Lim TH, Won You J, et al. A biomechanical comparison ofmodern anterior and posterior plate fixation of the cervical spine. Spine2001;26:15–21.

47. Wilke HJ, Claes L, Schmitt H, et al. A universal spine tester for in vitroexperiments with muscle force simulation. Eur Spine J 1994;3:91–7.

48. Panjabi MM. Dreidimensionale Testung der Stabilitat von Wirbelsaulenim-plantaten. Orthopade 1991;20:106–11.

49. Wilke HJ, Wenger K, Claes L. Testing criteria for spinal implants: recom-

mendations for the standardization of in vitro stability testing of spinalimplants. Eur Spine J 1998;7:148–54.

50. Pfeiffer M, Griss P, Franke P, et al. Degeneration model of the porcinelumbar motion segment: effects of various intradiscal procedures. Eur SpineJ 1994;3:8–16.

51. Gilad I, Nissan M. A study of vertebra and disc geometric relations of thehuman lumbar and cervical spine. Spine 1986;11:154–7.

52. Panjabi MM, Duranceau J, Goel V, et al. Cervical human vertebrae: quan-titative three-dimensional anatomy of the middle and lower regions. Spine1991;16:861–9.

53. Panjabi MM, Oxland T, Takata K, et al. Articular facets of the human spine:quantitative three-dimensional anatomy. Spine 1993;18:1298–310.

54. Penning L. Functional anatomy of joints and discs. In: The Cervical SpineResearch Society, ed. The Cervical Spine. Philadelphia: Lippincott, 1989.

55. Aerssens J, Boonen S, Lowet G, et al. Interspecies differences in bone com-position, density, and quality: potential implications for in vivo bone re-search. Endocrinology 1998;139:663–70.

56. Sikoryn TA, Hukins DW. Mechanism of failure of the ligamentum flavum ofthe spine during in vitro tensile tests. J Orthop Res 1990;8:586–91.

57. Jiang H, Moreau M, Raso VJ, et al. A comparison of spinal ligaments:differences between bipeds and quadrupeds. J Anat 1995;187:85–91.

58. Karaikovic EE, Daubs MD, Madsen RW, et al. Morphologic characteristicsof human cervical pedicles. Spine 1997;22:493–500.

59. Carson WL, Duffield RC, Arendt M, et al. Internal forces and moments intranspedicular spine instrumentation: the effect of pedicle screw angle andtransfixation: the 4R–4Bar linkage concept. Spine 1990;15:893–901.

60. Gaines RW Jr. Current concepts review: the use of pedicle-screw internalfixation for the operative treatment of spinal disorders. J Bone Joint Surg Am2000;82:1458–76.

61. Foley KT, DiAngelo DJ, Rampersaud YR, et al. The in vitro effects of instru-mentation on multilevel cervical strut-graft mechanics. Spine 1999;24:2366–76.

62. Wang JL, Panjabi MM, Isomi T. The role of bone graft force in stabilizing themultilevel anterior cervical spine plate system. Spine 2000;25:1649–54.


spine volume 30, number 11, pp 1275–1282 ©2005 ... · specimens and 20 mm for porcine specimens....

Documents