Journal (ir Spinal Disordl'rsV"I. 7. No. J. pp 19-28f "»)4 R::"cnPress,Lid..NewYork

Transpedicular Screw Fixation for Traumatic Lesions of theMiddle and Lower Cervical Spine: Description of the

Techniques and Preliminary Report

Kuniyoshi Aburni, *HajimeItoh, Hiroshi Taneichi, and Kiyoshi Kaneda

Department olOrthopaedic S!lrgery, Hokkaido University Schoo/ 01Medicine, Sapporo, and *Department 01Orlhopaedic S!lrgery, Kushiro Rosai Hospital, Kushiro, Japan

Summary: Thirteen patients with fractures and/or dislocations ofthe middle andlower cervical spine were treated by transpedicular screw fixation using the Steffeevariable screw placement system. Postoperative immobilization was either notused or simplified to short-tenn use of a soft neck collar. Recovery of nervefunction and correction ofkyphotic and/or translational defonnities were satisfac-tory. All patients had solid fusion without loss of correction at the latest follow-up.There were no neurovascular complications. It was concluded that transpedicularscrew fixation is as strong a fixation procedure for the cervical spine as it is for thethoraeie and lumbar spine. This surgical procedure is associated with some risks ofmajor neurovascular injuries; however, safety is adequate ifthe procedure is per-formed by experienced surgeons using meticulous surgical techniques. KeyWords: Transpedicular screw fixation-Cervical spine-Spinal injury.

The indications for surgical stabilization of cervicalspinal fractures and dislocations have been controver-sial.However, at the present time, in cases of neurolog-icai ;, v'olvementmost investigatorsrecommendsur-gicai reduction, decompression, and stabilization(3,I5,19,21). For the traumatic unstable cervicalspine, spinal instrumentation, e.g., anterior cervicalplate, spinous process wiring, Luque SSI, posteriorplate-screw fixation, and others are available for im-mediate stabilization of the unstable segment. In pos-terior screw fixation methods for the middle andlowercervical spine (i.e., the plate-screw fixation ini-tially performed by Roy-Camille and the hook-platestabilization by Mager! et al. and others), screws areinserted into the lateral portion of the articular

Address correspondence and reprint requests to Dr. KuniyoshiAbumi, Department ofOrthopaedic surgery, Hokkaido UniversityScl100lofMedicine, N-15 W-6, Kitaku, Sapporo 060, Japan.

This work was presented in part at the 66th Annual Meeting ofJapan Orthopaedic Association, Fukuoka, Japan, April 16-20,1992.

masses, not extending beyond the posterior column(5,9,13,19-21). Thus, there is not much difference instability between the posterior cervical plate and non-screw fixation methods based on the results of compar-ative biomechanical studies, with the exception oftheflexural stability provided by the posterior hook-platedescribed by Ulrich et a1. (4,8,23,24). On the otherhand, in the thoracic and lumbar spine, transpedicu-lar screw fixation methods offer three-column stabil-ity and have proven to be the most rigid posteriorfixation methods (12).

In the upper cervical spine, the procedure for poste-rior CI-2 transarticular screw fixation described byMager! et a1.does not call for the screws to be insertedinto the C2 pediele; they cross the isthmus elose to its,posterior surface and exit C2 at the posterior rim ofthe upper articular surface (14). Roy-Camille et a1.have reported direct screw fixation of C2 pedieles in ahangman fracture (20,21). However, there have beenno reports referring to transpedicular screw fixation inthe middle or lower cervical spine.

19

20 K. ABUMI ET AL.

Our concept was that screws for posterior plate fixa-tion inserted into the vertebral body through the pedi-cle would. provide better stability and strength thanthe othe'r fixation anchors, including the lateral massscrew in the cervical spine. In August 1990, the seniorauthor (K.Ao) started performing transpedicularscrew fixation with the Steffee variable screw place-ment (VSP) system (22) to treat middle and lowercervical spinal injuries. The purpose ofthis study is toreport the early results in 13 cases of middle or lowercervical spinal injuries treated by transpedicularscrew fixation, to describe our indications and tech-niques, and to recommend this procedure for moreextended cervical disorders in addition to traumaticlesions.

MATERIALS AND METHODS

Thirteen patients with middle or lower cervical spi-nal injury were treated by transpedicular screw fixa-tion using the Steffee VSP system at Kushiro RosaiHospitalandHokkaido University Hospital betweenAugust 1990 and April 1992 (Table 1).Therewere 12men and one woman, with an average age of 43.2years (range 15-80). The cause of the injury was a fallin six patients, a motor vehicle accident in five, and asports injury in two. The types of fractures and dislo-cations were described according to the mechanisticclassification of Allen et al. (2). Distractive flexion in-jury including stages 1, 2, and 3 occurred in sevencaSes. Stage 1 compressive extension injury occurredin five cases, .including four cases of unilateral supe-rior articular process fracture and one case of a separa-tion fracture ofthe articular mass described by Judetet al. (11). Stage 3 vertical compression injury oc-curred in one case. Impaired motion segments wereC4-5 in three patients, C5-6 in two, C6-~ in five,C7- Tl in one, C4-6 in one with stage 3 vertical com-pression injury, and C5-7 in one with aseparationfracture of the C6 articular mass. Preoperatively, ac-cording to the Frankel criteria (7), there were threepatients with complete loss of spinal cord function(grade A) and five patients with an incomplete spinalcord lesion (grade B in two, grade C in one, and gradeD in two). Another five patients with stage 1compres-sive extension injury had a single nerve root lesion.The time interval from injury to surgery ranged from7 h to 81 days.

Preoperative Management

Patients were immobilized with a Philadelphia col-lar. Cervical traction with tongs was not used becatise

J Spinal Disard Val. 7. No. I. 1994

,..;

E'"oe:.,.EU4;""c'"tt>..Dc

02Üc.:<;;U°iihoÖ...::>.,Z

c0-0Ü;~::> .,LI.-

LI. LI.<1: <1:

:- - - +~r-r-I-r-U"I~U"II-r-r-U"I~1 1 I 1 I I I 1 1 I I 1 I

U"I~~~U"I'<tU"l'<tr-~~'<t'<tUUUUUUUUUUUUU

~r-r-r-r-U"I"'U"II-r-r-U"I'"1 1.1 1 I 1 1 1 1 I 1 I 1

\0\0\0\0 V')v \f')v r---\O\Ov vUUUUUUUUUUUUU

.....oc

og"'~UNI.C ~"in :~'"vi)

.g 5",-'c<'".<:U.,~

~1-Nf'f")'VV')\Or---OOO\:::=~~U

co.;;;.2..ooe:uCca

TRANSPEDICULAR SCREW FIXATION

of magnetic resonance imaging (MRI) examination,except in one case of extremely unstable stage 3 dis-tractive flexion injury. Plain roentgenography, antero-posterior (AP) and lateral laminography, computedtOmography (CT), and MRI were performed in allcases. Myelography was also performed in six cases.

Implants

The Steffee VSP system was originally~designed forthe lumbar spine and then used in the thoracic spine,so even the smallest plate is too large for the cervicalspine. To fix a single motion segment in the middle or(owercervical spine, an original plate with two nestedslots was divided, and both ends of the plate wereshaved off with a grinder to avoid irritating the adja-cent facet joints. Tbe tips of 4.5-mm diameter VSPsrrcws were cut off to proper length before or duringsurgery.

SurgicaI Procedure

After intubation, patients were put into the proneposition on a Relton-Hall frame. The head was tapedto the headrest with the cervical spine maintained in aneutral position, and the shoulders were pulled cau-dally by a heavy bandage for intraoperative lateral ra-diographie imaging ofthe lower cervieal spine. A pos-terior midline skin incisiön was made, and the para-vertebral museles were dissected latera1ly to exposethe lateral margins of the compromised facet joints.Because the insertion angle ofthe screw was intendedto be 30-400 medially to the midline in the transverseplane, a longer skin incision was required than for astandard spinous process wiring procedure. In the sixpatients with facet-interlocking, stage 2 and 3 distrac-tive !lexion injuries, partial resection ofthe facetjointwas performed bilaterally or unilaterally for reduc-tion. In the flve cases of nerve root compression byarticular bone fragment, the fragment in the neuralforamen was removed after partial resection of theinferior articular process.

The point of screw penetration at the posterior cor-tex of the articular mass was determined slightly lat-eral to the center of the artieular mass and elose to thePOSl<:riormargin ofthe superior artieular surface, tak-ing into consideration. the location of the vertebralartery, the spinal cord, and the pediele (Fig. IA). Be-fore inserting the screws, the cortex at the point ofinsertion was penetrated with a high~speed burr andthe entrance hole was enlarged to proper size to burythe screw anchor into the articular mass. After creat-ing the screw insertion hole, the entrance ofthe pedi-eIewas visible directly (Fig. 2A). A small nerve retrac-

21

B

FIG. 1. A: Landmarks for placement of screw insertion holes(arrow). B: Direction of the screw showing its relationship to thearticular facet, vertebral artery, and spinal canal.

tor was used for a pedicle probe, and the inner wall ofthe pediele cavity was palpable with the retractor. Thedirection and the insertion depth ofthe retractor andthe screw were confirmed by the intraoperative lateralimage intensifier. In the patients with facet interlock-ing and those with artieular process fractures, the in-feromedial portion of the upper pedicle and the su-peromedial portion of the lower pedicle were palpablewith the nerve retractor through the defect created byexcision of the articular processes. Laminotomy wasnot performed to identify the medial aspect of thepedicle.

The intended angle ofthe screws based on measure-ments of preoperative CT images was 30-400 medialto the midline in the transverse plane (Fig. IB), andparallel to the upper end-plate in the sagittal plane.Drilling of the pedicle was not performed to avoidneurovascular injuries. Insertion of the screw wasgreater than two thirds of the AP vertebral bodydepth, and the anterior cortex was penetrated by fourscrews in two cases.

MRI demonstrated retropulsed disk materials atthe injured level in two patients. In those patients, toavoid neurological deterioration, slight distractionforce was applied between the upper and lower screwsbefore tightening the nuts. Decortication of the lam-ina and spinous processes was performed after com-pietion of the instrumentation. Finally, cmppedbones were routinely placed between the spinous pro-cesses and on the remaining exposed facet joint andlaminae. An H-shaped bone was grafted between thespinous processes in six patients (Fig. 2B). The graftedbone was obtained from the ilium in nine cases andfrom the adjacent spinous processes in four. One-Ievel

J Spinal Disord Val. 7. No. 1. 1994

22 K. ABUMI ET AL.

fixation was performed in 10 patients, whereas two-level fixation was required in three patients, i.e., in astage 3 vertical compression injury, in a C6-7 com-pressive extension injury associated with disruptionofthe C7 vertebral body, and in aseparation fractureofthe C6articular masswith instability at two lp.otionsegments.In the case of the vertical compression in-jury with severespinal cord compressionby a commi-nuted vertebral body, additional anterior decompres-sion and an iliac strut bone graft was performed afterthe posterior surgery. In one of the casesof compres-sive extension injury, one of the pedicles was dis-tractedby screwing,and plating wasperformedunilat-erallywith subsequent anterior interbody fusion. Allsurgery was performed by the senior author (K.A.),except in one case in which it was performed by thesecond author (H.I.), and the operative techniquesweresame in all cases.

Postoperative Management and Qinical Evaluation

Postoperative immobilization varied according tothe patient's neurological status, general condition,and type of employment. Patients with severespinalcord lesions(Frankel grade'A or B)started their reha-bilitation 5-10 days after surgerywithout any exter-nal support. Eight patients with mild nerve lesions(Frankel grades C and D, and radiculopathy) wereallowedto ambulate with a soft neck collar the dayafter surgery. These eight patients and one of theFrankel B patients returned to their originaljobs 3-6weeksafter surgery,before bony union wascomplete.The collar was worn for 8 weeks in fivepatients whoreturned to heavy work. However, the collar was

J Spinal Disard Val. 7. No. /. /994

FIG. 2; A: Direct exposure of thepedicle cavity in the screw inser-tion hole (arrow). B: Completion ofinstrumentation arid H-shapedbane graft (open arrow) betweenthe spinous processes.

worn only for 2-3 weeks in three patients who re-turned to light work.

Follow-upphysical examinations and radiographieevaluations were performed on all patients.To assessstability and deformity, flexion-extension andoblique x-ray films were obtained in all patients 3, 6,9, and 12 months after surgery,and ifpossible at 18and 24 months. Laterallaminograms were obtainedin all patients 6 and 12 months after surgery.Bonyunion was evaluated independently by three of theauthors (K.A., H.I., H.T.). An assessment of solidunion was made on the basis of the presen~eof a ho-mogeneous fusion mass on laterallaminograms and asegmental motion of <2° on flexion-extensionfilms.Postoperative CT scans were obtained in all patientsto assessscrewplacement. The CT windowand widthwere adjusted to reduce metal artifacts. The range offollow-up for these patients was 12-32 months, withan averageof 22 months.

RESULTS

Duration of surgery and intraoperative blood lossfor the posterior procedure were measured, excludingthe anterior surgical portion, in two cases of com-bined anterior and posterior surgery. The averageoperation time was 145 min (range 95-210), and theaverage intraoperative blood loss was 283 ml (range50-600). There were no increases in neurologie deficitin this series. Using the Frankel criteria, three FrankelA patients remained Frankel A; however, two ofthesethree patient showed descent of the level of paralysisby one or two spinal cord segments. Two Frankel B

TRANSPEDICULAR SCREW FIXATION 23

paticnts improved to Frankel C and D, and one Fran-kel C patient, two Frankel D patients, and five pa-tients with radiculopathy returned to normal neuro-logie status. .

At the latest follow-up, all cases had solid bonyunion. Three judges were unanimous about bonyunion in all cases. There were no cases of im plantcomponent, connection, or bone interface faHure. Ky-photic and translational deformities were satisfacto-rily ;:orrected, and the correction was maintained in

FIG.3. Case 7. Distractive flexioninjury(stage2) with ca radiculopathy on the left and a mildspinal cord lesion. A-C: Lateral roentgenogramshows C5-6 subluxation. LaterallaminogramandCT show unilateralfacet interlocking.D-G: Afterreduction of the interlockingand decompressionof the nerve root, C5-6 is stabilized. Postopera-tive CT demonstrates proper direction of thescrews. Ridges of the right screw threads seemto penetrate the medialcortex of the rightca pedi-eie. Neurologicalrecovery was complete.

all cases. No patients had loss of correction ofkypho-sis > I 0; and correction oftranslation was maintainedcompletely in all cases. The averages and ranges oftranslational and kyphotic deformity before surgery,immediately after surgery, and at follow-up are shownin Table 2. There were no complications involving thespinal cord, nerve roots, or vertebral artery.

Assessment of screw placement into die pedicle bypostoperative CT images was difficult because of themetal artifacts, but the direction of the inserted screw

J Spinal Disard Val. 7. No. I. 1994

24 K. ABUMI ET AL.

could be assessed. The angle of the inserted screwsranged from 25° to 45° medial to the midline in thetransverse plane. The medial cortex ofthe three pedi-eIes in two patients, two of the C5 pedieIes and one ofthe C6 pedieIes (Fig. 3), appeared to be perforated bythe screw threads. In one other case, according to theoblique film assessment, the site of insertion of onescrew into the C5 pedieIe was too far inferior, andsome ofthe screw threads had obviously intruded intothe neural foramen. There 'Yere no neurological com-plications in these three cases.

CASE PRESENTATIONS

The following are the case presentations (Table I)with their corresponding figures: case 7 (Fig. 3), case12 (Fig. 4), case 13 (Fig. 5), and case 11 (Fig. 6).

J Spinal Disord Val. 7, No. /, /994

FIG. 4. Case 12. Distractive flexion injury(stage 3) with complete spinal cord lesion. Aand B: Lateral roentgenogram shows bilat-eral facet dislocation at C4-5. C and D: Post-operative roentgenograms demonstrate re-alignment of the spine.

DISCUSSION

Surgical decompression and stabilization is re-quired to treat most of the cases of cervical spinalinjuries with neurological impairment. The decom-pression and stabilization procedure should differde-pending on the pathologyin each case.Surgicalcorrec-tion ofthe malalignment and stabilization by the iso-lated posterior or the combined anterior and posteriorapproach is common in casesof compressiveflexion,compressive extension, distraction flexion, and stage2 distractive extension injuries (15,19-21). Con-versely,in vertical compression injuries with anteriorspinal cord compression, anterior decompressionandstabilization are indicated.

Usually, some ofthe anterior or posterior stabiliz-ing structures are preserved to some extent, except in

TRANSPEDICULAR SCREW FIXATION 25

FIG.5. Case 13. Separation fracture of the C6 articular mass on the right side with radiculopathy.A-C: Lateral roentgenogram shows slight anterior translation of C6. Oblique laminogram and CTdemonstrate separation fracture of the C6 articular mass and narrowing of the right C5-6 interver-tebral foramen. D-G: After decompression of the C6 nerve root, C5-7 were stabilized. Postopera-tive films demonstrate proper screw placement.

stage 4 distractive flexion injuries or stage 5 compres-siveextension injuries. Therefore, from a biomechan-ical view point, it is preferable to avoid surgical de-struction ofthe residual stabilizing structures. For thisreason, an optimal procedure should pe a single ante-rior or posterior method augmented by rigid instru-mentation. Generally, posterior fixation devices havean advantage over anterior devices for the fixation ofposterior injuries. Similarly, anterior fixation devicesare more suitable than posterior devices for the fixa-tion of anterior injuries (1).

S;)inal fixation devices are required to control all

types of spinal motion, i.e., translation along and ro-tation about each of the three axes of spinal motion.In addition, this multidirectional stabilizing capabil-ity is preferablyobtained by isolatedanterior or poste-rior short segmental fixation, savingas many freemo-bile segments as possible. However, the cortex of thevertebral bodies in the cervicalspine isnot as strongasthat of the posterior bony elements. Consequently,even ifthe screwspenetrate the posterior cortex ofthevertebral body, anterior plate fixationsda not provideas sufficientstability as posterior fixation proceduresin cases of posterior injury or combined anterior and

J Spinal Disard Val. 7, No. 1. 1994

26 K. ABUMI ET AL.

posterior injuries (4,23,24). Additional use of poste-rior instrumentation on anterior cervical plates maybe recommended. At the present time, spinous pro-cess wiring, posterior plating, Luque SSI, and othertechniques are available to stabilize the unstable cer-vical spine posteriorly. However, based on biome-chanical studies, none of these fixation devices pro-videsadequate stability in the combined anterior andposterior discoligamentouscervical injury model, ex-cept for flexuralstability results provided by posteriorhook-plate (4,8,23,24).Accordingto the resultsofthebiomechanical studies of posterior plate fixation byGill et al. (8) and Montesano et al. (16), bicorticalscrewsprovideda better stabilizingeffectthan did uni-

J Spinal Disord Val. 7, No. 1, 1994

FIG. 6. Gase 11. Vertical compression injury (stage 3) of GSwith spinal cord lesion, Frankel B. A-G: Lateral roentgeno-gram and GT show spinal canal encroachment by a retro-pulsed bony fragment. MRI demonstrates spinal cord com-pression by the fragment. D-G: Posterior stabilization afteranterior decompression and iliac bone graft. Lateral and APfilms demonstrate realignment of the spinal canal. Postopera-tive GT shows anterior decompression and proper insertion ofthe screws through the pedicles.

cortical screws.However, the stiffnessprovided by bi-cortical screwsdid not exceed that produced by spi-nous process wiring. Furthermore, screw insertioninto the lateral portion of the articular mass exposesthe spinal nervesor vertebral artery to injury (10),andsublaminar wiring places the cervical spinal cord atrisk, especially in cases of narrowed spinal canal.Based on biomechanical studies, among establishedstabilizingprocedures, only the combination of ante-rior and posterior instrumentation can beexpected toprovide sufficientstability for an extremely unstablecervical spine without major extemal support (24).

Transpedicular screw fixation systems have beendeveloped for the thoracic and lumbar spine, and

TRANSPEDICULAR SCREW FIXATION

Note. Negative numbers in parentheses refer to lordosis. A case ofverticaJ compression injury of C5 was exduded from measure-ments oftranslation.

their excellent stability and strength allows the pa-!ients to walk immediately after surgery without anyextanal support. Despite the perceived risks of tho-racic and lumbar transpedicular screw placement, ex-act knowledge ofthe anatomy ofthe pedicle and care-ful surgical techniques allow safe placement. Exceptfor pedicular screwing of C2 by Roy-Camille (20,21),transpedicular screw fixation in the cervical spine hasnot been performed for fear of injuring the vertebralartery, spinal cord, and nerve roots. Roy-Camille~'.at,:dthat screwing into the C3 to C6 pedicle wouldbe an unacceptable risk (21). However, the pedicle ofthe cervical spine is a strong structure ofthe vertebra,as in the thoracic and lumbar spine. Thus, the stabil-ity obtained by transpedicular screw fixation in thecervical spine should be equal to or greater than thatachieved by combined anterior and posterior instru-mentation. However, further comparative biome-chanical study is needed to assess the stabilizing effectb.: '-~:isprocedure precisely.

The most severe complication of transpedicularscrew fixation of the cervical spine is injury of thevertebral artery or spinal cord. Such complications,which could be fatal, should be completely preventedby exact screw placement into the pedicles. In somepatients, especiaIly in females, the diameter ofthe ped-icles is too smaIl to insert a screw. Consequently, CTexamination, adjusted to the bony elements, is recom-me~ldedpreoperatively to determine the diameter ofthe pedicles, and the patients with very smaIl pediclediameters should be excluded from this procedure.Direct exposurt:; of the pedicle cavity by creation of ahole at the insertion points and the use of pedicleprobe and image intensifier are essential to the safetyand success of cervical pedicle screw placement. Inaddition, we recommend confirmation of the loca-tion of the medial, superior, and inferior surfaces ofthe ;..;~dicleswith a small nerve retractor, if possible, in

27

cases in which facetectomy andjor laminotomy hasbeen performed.

Panjabi et al. have reported a detailed anatomicalstudy of the pedicles of the cervical spine (17). Ac-cording to them, the averageangle ofthe pedicleaxesfrom lateral to midline in the transverseplane was45°at C4, 39° at C5, 29° at C6, and 33° at C7. H.owever,the screw insertion angle in this series, slightly lessthan these values in some cases, was acceptable andsafe enough for both the spinal cord and the vertebralartery. Becauseofthe small depth ofthe pediclein thecervical spine, the direction of the screw insertion isnot so severelyrestricted. The cervicalnerve roots ronanterolaterally at -45° with respect to the coronalplane and downward "'":'10° with respect to the trans-verse plane (6,18). Within the foramina, they are 10-cated at and below the disk level, i.e., in the inferiorhalf of the neural foramina (18). Thus, there is someroom between the medial and inferior surface of thepedicle and the neural elements.Slight perforation ofpedicleby screwthreads in the medialor inferior direc-tion is relatively safe for the spinal cord and nerveroots in the cervical spine. Anterior to the vertebralbodies covered by the anterior longitudinal ligament,the pharynx is located above C4 and the esophagusbelow C5 in the median portion. If screw insertion istoo deep, the constrictor of the pharynx or esophagusmay be injured. Too deep screw insertion, such aSmore than a few millimeter beyond the vertebralbody, should be avoided.

In this series, perforation of the pedicle by screwthreads was suspected in three pediclesand was obvi-ous in one pedicle, despite proper screw direction.The cause of these perforations is suspected to bemainly the maladaptation betweenscrew and pediclediameter. Accordingto the valuesreported by Panjabiet al., pedicle diameter is smallest.at C3 or C4 andincreases in sizetoward the cervicothoracicjunction.Thus, in the mid-cervical spine, even the screw withthe smallest diameter of 4.5 mm is sometimes toolarge. Screwsof appropriate size for the mid-cervicalspine need to be made.

Transpedicular screwfixation,with its stability andstrength, appears to be a useful stabilizing procedurefor the reconstruction of the injured cervical spinewithout destruction of the pedicle or vertebral bodysuch as distractive flexion or compressive extensioninjuries. Moreover,this procedureisapplicableaspos-terior reinforcement for vertebral body replacementin casesof vertical compression injury, spinal tumor,or tuberculous spondylitisofthe cervicalspine. In ad-dition to these uses, transpedicular screw fixationwould be useful for longer fixation,e.g., fixation from

J Spinal Disord Val. 7, No. I. 1994

TABLE 2. Angular and translational deformityin the sagiual plane

AtPreoperative Postoperative follow-up

Kyphosis(n = 13)10.9° (-0.5°) 0°Average

Range 3-21° (_4°)_3° (-3°)_2°Translation(mm)

(n =12)Average 6.4 0.3 0.3Range 1-14 0-2 0-1

28 K. ABUMI ET AL.

the upper to lowercervical spine or from the middlecervical spine to the thoracic spine. It would also beuseful for casesof destruction of the lamina becausein such"atientsit is difficultto stabilize the spine byspinous processwiring or sublaminar wiring withoutsacrificingthe adjacent intact motion segments.

From a reviewofthis study, the stiffinternal stabi-lizingeffectof transpedicular screwfixation precludesthe peedforadditionalanterior surgeryand postopera-tive external support evenfor extremelyunstable cer-vical spinessuchas in the four casesof stage 3distrac-tive flexioninjurytreated in this series.The procedureallowedeasynursingcare, early ambulation and reha-bilitation, and early return to the patients previousjob. There were no pseudarthroses or complicationsinvolving the vertebral artery, spinal cord, or nerveroots. Direct exposure of the pedicle cavity beforescrewplacementand the help of an image intensifieradequately confinned screwinsertion. However, thissurgical procedure is associated with some risks ofmajor neurovascular injuries. Therefore, it requiresprecise knowledge of the anatomy of the cervicalspine and meticuloussurgicaltechniques, and shouldbe performed only by the surgeonsexperienced bothin transpedicular screw fixation in the thoracic andlumbar regionsand in surgery of the cervicalspine.

CONCLUSIONS

Transpedicular screw fixation with the Steffee VSP-system provided high stability and strength in thetreatment oftraumatic middle and lower cervical spi-nallesions without any neurovascular complications.However, this procedure is associated with some risksof major neurovascular injuries. It should be per-fonned only by experienced surgeons using the metic-ulous surgical techniques described. In addition totraumatic lesions, we also recommend this procedurefor more extended cervical reconstructions.

REFERENCES

1. Abumi K, Panjabi MM, Duranceau J: Biomechanical evalua-tion of spinal fixation devices: part 1II. Stability provided by sixspinal fixation devices and interbody bone graft. Spine14:1249-1255,1989

2. Allen BL, Ferguson RL, Lehmann TR, O'Brien RP: A mecha-nistic classification of closed, indirect fractures and disloca-tions ofthe lowercervicalspine.Spine 7:1-27, 1982

3. Bohlman HH: Acute fractures and dislocations ofthe cervicalspine. J Bone Joint Surg [Am] 61:1119-1 ]42, 1979

4. Coe JD, Warden KE, Sutterlin CE IlI, McAfee PC: Biome-chanical evaluation of cervical stabilization methods in a hu-man cadaveric model. Spine 14:1122-1131,1989

J Spinal Disord Val. 7. No, 1. 1994

5. Cooper PR, Cohen A, Rosiello A, Koslow M: Posterior stabili-zation of the cervical spine fraclUres and subluxations usingplates and screws. NeuroSllrgery 23:300-306, 1988

6. Danie]s DL, Hyde JS, Kneeland JB, et al.: The cervical nervesand foramina: local-coil MRI imaging. AJNR 7:129-133, 1986

7. Frankel HL, Hancock DO, Hyslop G, et al.: The value ofpos-tural reduction in the initial management of closed injuries ofthe spine With paraplegia and tetraplegia. Paraplegia 7:179-192, 1969

8. GilI K, Paschal S, Corin J, Ashman R, Bucholz RW: Posteriorplating ofthe cervical spine. Spine 13:813-816, 1988

9. Heller JG, Carson GD, Abitbol JJ, Garfin SR: Anatomie com-parison of the Roy-Camille and Magerl techniques for screwplacement in the lower cervical spine. Spine 16:S552-557,1991

10. Jonsson H Jr, Rauschning W: Postoperative cervical spine spec-imens studied With the cryoplaning technique. J OrlhopTrauma 6:1-9, 1992

11. Judet R, Roy-Camille R, Zerah JC, Saillant G: Fracture-separa-tion du massif articulaire. Rev Chir Or/hop 56: 155-164, 1970

12. Krag MH: Biomechanics of transpedicular screw fixation. In:The Lumbar Spine, ed by Weinstein JN, Wiesel SW, Philadel-phia, WB Saunders, 1990, pp 916-940

13. Magerl F, Grob D, Seemann P: Stable dorsal fusion ofthe cer-vical spine (C2-Tb I) using hook plates. In: Cervical Spine I. edby Kehr P, Weidner A, New York, Springer-Verlag, 1987, pp217-221

14. Magerl F, Seeman P: Stable posterior fusion of the atlas andaxis by transarticular screw fixation. In: Cervical Spine I, ed byKehr P, Weidner A, New York, Springer-Verlag, 1987, pp322-327

15. McAfee PC, Bohlman HH: One-stage anterior cervical decom-pression and posterior stabilization With circumferential ar-throdesis.J BoneJoint Surg [Am] 71:78-88, 1989

16. Montesano PX, Jauch E, Jonsson H Jr: Anatomie and biome-chanical study of posterior cervical spine plate' arthrodesis: anevaluation of two different techniques of screw placement. JSpinal Disorders 5:301-305, 1992

17. Panjabi MM, Duranceau J, Goel VK, Oxland T, Takata K:Cervical human vertebrae. Quantitative three-dimensionalanatomy ofthe middle and lower regions. Spine 16:861-869,1991

18. Pech P, Daniels DL, Williams AL, Haughton VM: Tbe cervicalneural foramina: correlation of microtomy and CT anatomy.Radiology 155:143-146, 1985

19. Roy-Camille R, Saillant G, Berteaux, et al.: Early managementof spinal injuries. Recent advances in orthopedics, ed byMckibbon B, Edinburgh, ScotIand, Churchill Livingstone.1979, pp 57-87

20. Roy-Camille R, Saillant G, Mazel C: Internal fixation of theunstable cervical spine by a posrerior osteosynthesis with platesand screws. In: The Cen'ical Spine, 2nd ed, ed by Tbe CervicalSpine Research Society, Philadelphia, 18 Lippincott, 1989, pp390-403

21. Roy-Camille R, Mazel C, Saillant G, Benazet JP: Rationaleand techniques of internal fixation in trauma of the cervica1spine. In: Spinal Trauma. ed by Errico T, Bauer RD, Waugh T,Philadelphia, JB Lippincott, 1991, pp 163-191

22. Steffee AD, Biscup RS, Sitkowski DJ: Segmental spinal platesWith pedicle screw fixation. A new internal fixation device fordisorders ofthe lumbar and thoracolumbar spine. Clin Or/hop203:45-53, 1986

23: Sutterlin CE, McAfee PC, Warden KE, Rey RM, Farey 10: Abiomechanical evaluation of cervical spinal stabilization meth-ods in a bovine model. Spine 13:795~802, 1988 .

24. Ulrich C, WoersdoerferO, KalffR, Claes L, WilkeH-J: Biome-chanics offixation systems to the cervical spine. Spine 16:54-S9,I991