Antonio Leone, MDAlfonso Cerase, MD2

Cesare Colosimo, MDLuigi Lauro, MDAlfredo Puca, MDPasquale Marano, MD

Index terms:ReviewSkull, base, 127.41Skull, fractures, 127.41Trauma, 127.41

Radiology 2000; 216:635–644

Abbreviations:CCJ 5 craniocervical junctionOCF 5 occipital condylar fracture

1 From the Departments of Radiology(A.L., A.C., C.C, L.L., P.M.) and Neu-rosurgery (A.P.), Universita Cattolicadel Sacro Cuore, Policlinico “AgostinoGemelli,” Largo Agostino Gemelli, 8,00168 Rome, Italy. Received January12, 1999; revision requested March 5;final revision received September 13;accepted September 14. Address cor-respondence to A.L. (e-mail:a.leonemd@tiscalinet.it).

Current address:2 Department of Neuroradiology, Azien-da Ospedaliera Senese, Policlinico “LeScotte,” Siena, Italy.

© RSNA, 2000

Occipital Condylar Fractures:A Review1

The purpose of this review article is to summarize the epidemiology, pertinentanatomy, mechanisms of injury, and classification systems of occipital condylarfractures (OCFs), as well as their clinical presentation and screening, the importanceof computed tomography (CT) for detection, and current treatment options. Theauthors emphasize the rate of occurrence of OCFs, which may be detected in asmany as 16% of patients with craniocervical injury. Clinical presentation is notspecific, and OCF is not readily diagnosed at physical examination. Failure todiagnose may result in substantial morbidity, and thus accurate diagnosis is man-datory for both therapeutic and medicolegal implications. The diagnosis is mostlikely to be made with CT. Thin-section CT technique is the method of choice toevaluate the traumatized craniocervical junction. OCFs should be suspected in allpatients sustaining high-energy blunt trauma to the head and/or upper cervicalspine, resulting from axial loading, lateral bending and/or rotation, and/or directblow. Besides a CT study assessing potential intracranial injuries, these patientsrequire CT of the craniocervical junction. Radiologists should be aware of the typesof OCFs and associated injuries.

Traditionally thought of as rare, occipital condylar fractures (OCFs) have recently beenconsidered as an underdiagnosed condition that likely occurs with greater frequency thanis generally accepted (1). OCFs can easily be missed because the clinical manifestation ishighly variable and the results of physical examination are usually nonspecific (2–4);however, long-term morbidity due to pain and limited motion, serious neurologic deficits,or even death may result from undiagnosed or untreated OCFs (5). Thus, an accuratediagnosis is mandatory for both therapeutic and medicolegal implications. OCFs aretraumatic lesions of the skull base that are rather frequently, but not necessarily, associatedwith severe head, brain, and cervical spine trauma resulting from high-speed decelerationinsults (6–8). They are unusual clinical injuries because of the strategic anatomic locationof the occipital condyles within the craniocervical junction (CCJ) (3,9–11). Familiaritywith the types of OCFs, as well as their mechanisms of injury and clinical manifestation,is essential for radiologists evaluating patients with craniocervical trauma. As treatmentoptions change, precise identification of the type and extent of OCF, as well as ofconcomitant lesions of the CCJ, is becoming increasingly important in determiningappropriate management (5).

EPIDEMIOLOGY

In the past, OCFs were considered to be extremely rare because of their difficult detectionwith conventional radiography. To our knowledge, the first case was identified in 1817 bySir Charles Bell at the autopsy study of a victim of a fall (12); the second case was describedin 1900 (13). The first radiographic evidence of an OCF in vivo was reported in 1962 (14).The first computed tomographic (CT) scans of OCFs were published in 1983 (15,16). Sincethen, the widespread use of CT and the improvements in its technology, as well as bettertrauma care, have resulted in an increase in OCF reports in the literature (17–23). However,the true frequency of OCFs is still unknown. In a postmortem radiologic examination of312 victims of traffic accidents (24), an OCF was reported in two of the 186 patients withhead and/or neck injury. In a postmortem analysis of 112 consecutive victims of fatalmotor vehicle accidents (25), an OCF was noted in two of the 26 patients with cervicalspine injuries. From another postmortem study of 155 persons killed in traffic accidents,

Review

635

OCFs were seen in three of the 66 pa-tients with trauma to the cervical spineor CCJ (26–28). Other authors (29) de-scribed the occurrence of OCFs in 25 of600 patients who died in traffic accidentsbut did not report the number of patientswith craniocervical injuries. As regardsthe CT evaluation of patients survivingtheir craniocervical trauma, one study(30) of patients with severe head injury(defined as major trauma resulting in asubstantially altered mental status and aGlasgow Coma Scale score of 3–6 on ad-mission) reported a 4% incidence ofOCFs. Another study (31) that includedpatients sustaining severe nonpenetrat-ing cervical trauma with a mean GlasgowComa Scale score of 10.8 (range, 3–15)reported a 3% incidence of OCFs. In amore recent study (1) that expanded theinclusion criteria to incorporate all pa-tients with appropriate mechanisms ofinjury (ie, high-energy blunt trauma tothe head and neck involving compo-nents of either axial compression, lateralbending or rotation, or direct blow), andirrespective of Glasgow Coma Scale score,the resultant incidence of OCFs was 16%(nine of 55).

ANATOMY

The occipital condyles are the promi-nences of the paired lateral exoccipitalsegments of the occipital bone, whichform the foramen magnum together withthe basioccipital segment anteriorly andthe supraoccipital or squamosal segmentposteriorly (32,33). The bone around theforamen magnum constitutes the upper-most border of an extremely complex

three-unit joint with intricate functionalrelationships between the occiput, atlas,and axis (ie, the CCJ or occipitoatlanto-axial complex) (33–38). The CCJ includessix synovial-lined articulations: the pairedoccipitoatloid joints, the anterior and pos-terior median atlanto-odontoid joints, andthe paired atlantoaxial joints. The mostcommon shape of the occipital condyle isoval or beanlike; it slopes inferiorly fromlateral to medial in the coronal plane andmakes an angle with the midsagittal planeof 25°–28° in adults (35) (Fig 1). The occipi-toatloid articulations are “cup-shaped”paired joints between the convex surfacesof the occipital condyles and the concavesuperior surfaces of the articular facets ofthe atlas. In the coronal plane, both theoccipital condyles and the superior articu-lar facets of the atlas slope downward me-dially. The anterior atlanto-odontoid artic-ulation lies between the anterior arch ofthe atlas and the anterior aspect of theodontoid process of the axis. The posterioratlanto-odontoid articulation is betweenthe posterior aspect of the odontoid pro-cess of the axis and the anterior cartilagi-nous aspect of the transverse portion of thecruciform ligament (ie, the transverse liga-ment of the atlas). The atlantoaxial articu-lations are paired joints between the infe-rior articular facets of the atlas and thesuperior articular facets of the axis (Fig 1).

Stability of the CCJ is provided by anumber of ligamentous structures thatcan be divided into two groups accordingto their attachments. The anterior longi-tudinal ligament, cruciform ligament,tectorial membrane, and nuchal liga-ment attach to all three bones. The ante-rior occipitoatloid membrane, atlanto-

odontoid ligament, apical ligament ofthe dens, alar ligaments, posterior occipi-toatloid membrane, and the atlantoaxialmembrane attach to two bones each (Fig1). The anterior longitudinal ligament at-taches to the anterior body of the axis,anterior arch of the atlas, and anteriorinferior edge of the occipital bone afterrunning the entire length of the spine. Inthe upper cervical spine, the anteriorlongitudinal ligament appears as a thin,translucent structure. The cruciform liga-ment has transverse and vertical por-tions. The transverse portion is the majorone and is most commonly known as thetransverse ligament of the atlas. It ex-tends between osseous tubercles on themedial aspects of the lateral masses of theatlas and consists almost exclusively ofcollagen fibers (38). Vertical portions in-clude an ascending band, attached to theanterior edge of the foramen magnum,and a descending band, attached to theposterior aspect of the body of the axis.The tectorial membrane is a broad andfairly strong band that is regarded as thecephalic extension of the posterior longi-tudinal ligament running from the pos-terior surface of the body and odontoidprocess of the axis to the anterolateraledge of the foramen magnum. It is lo-cated between the cruciform ligamentand the atlas anteriorly and the anteriordura mater posteriorly. The nuchal liga-ment, which extends from the posteriorborder of the occiput to the spinous pro-cess of C7, is attached to the spinousprocesses of the cervical vertebrae andthe interspinous ligaments. The anterioroccipitoatloid membrane runs from thecephalad portion of the anterior arch of

Figure 1. Schematics show the anatomy of the CCJ. A, Midsagittal view. B, Posterior view of coronal section passing through the occipital condylesand posterior arches of the atlas and axis after dissection of the dura mater, posterior longitudinal ligament, and tectorial membrane. C, Transverseview from above of the median atlanto-odontoid joints after dissection of the occiput, dura mater, posterior longitudinal ligament, tectorialmembrane, and anterior and posterior occipitoatloid membrane. abCL 5 ascending band of the cruciform ligament, AL 5 alar ligaments (atlantaland occipital portions), AOAM 5 anterior occipitoatloid membrane, AOL 5 atlanto-odontoid ligament, ApL 5 apical ligament, dbCL 5 descendingband of the cruciform ligament, DM 5 dura mater, OC 5 occipital condyle, PLL 5 posterior longitudinal ligament, POAM 5 posterioroccipitoatloid membrane, SAFA 5 superior articular facet of the atlas, TGVA 5 transverse groove of the atlas for vertebral artery, TL 5 transverseligament of the atlas, TM 5 tectorial membrane.

636 z Radiology z September 2000 Leone et al

the atlas to the anterior edge of the occi-put and is considered to be part of theanterior longitudinal ligament. The at-lanto-odontoid ligament runs betweenthe anterior surface of the odontoid pro-cess of the axis to the caudal portion ofthe anterior arch of the atlas.

The alar ligaments are paired structuresthat arise from the dorsolateral aspect ofthe odontoid process of the axis and runobliquely to connect with the inferome-dial aspect of the occipital condyles andthe lateral masses of the atlas. They con-sist mainly of collagen fibers, although afew elastic fibers may be identified in themarginal regions (38). The apical liga-ment of the dens connects the apex ofthe odontoid process of the axis with theanterior edge of the foramen magnum. Itlies between the ascending band of thecruciform ligament and the anterior oc-cipitoatloid membrane. The posterior oc-cipitoatloid membrane attaches to theposterior margin of the foramen mag-num and to the posterior arch of theatlas, while the posterior atlantoaxialmembrane runs between the posteriorarches of the atlas and axis. The basioc-ciput and the occipital condyles togetherform the attachments for the paired alarligaments, the apical ligament of thedens, and the ascending band of the cru-ciform ligament.

The clinical importance of OCFs is dueto the proximity of the occipital condylesto the medulla oblongata, vertebral arter-ies, and lower cranial nerves (33,39–43).The medulla oblongata, meninges, verte-

bral arteries, anterior and posterior spinalarteries, and veins that communicatewith the internal vertebral venous plexuspass through the foramen magnum,which is bound by the previously de-scribed segments of the occipital bone. Inparticular, the configuration of the atlan-toaxial segment of the vertebral arterynormally allows about 35° of head andatlas rotation. At the upper atlantal sur-face, the vertebral artery curves posteri-orly to its transverse groove on the atlas,behind the superior atlantal articularfacet. The vertebral artery usually doesnot contact the groove directly but is sep-arated by the vertebral venous plexus.The vertebral artery enters the subarach-noid space by piercing the posterior oc-cipitoatloid membrane and dura materjust medial to the occipital condyle.Within the subarachnoid space, the ver-tebral artery takes either a straight or (es-pecially in the elderly) a curved patharound the lateral and anterior aspects ofthe spinal cord and medulla oblongata tomerge with its counterpart at the lowerend of the pons to form the basilar artery(Fig 2a) (39).

Within the base of each occipital con-dyle lie the hypoglossal (anterior condy-loid) canals through which pass the hy-poglossal nerve (cranial nerve XII), ameningeal branch of the ascending pha-ryngeal artery, and an emissary vein (Fig2b, 2c) (33,43). Lateral to the occipitalcondyle and hypoglossal canal and pos-terior to the carotid canal is the jugularforamen (posterior foramen lacerum),

which is a true canal containing the cra-nial nerves IX–XI, inferior petrosal sinus,internal jugular vein, and posterior men-ingeal artery. It lies inferolaterally withinthe temporal bone, between its petroussegment, anterolaterally, and the occipi-tal bone, posteromedially (Fig 2b, 2c).The variability in bone formation aroundthe primitive posterior foramen lacerum,the unequal development of the lateralsinuses, and the complex anatomic rela-tionships of the neurovascular structureswith each other result in asymmetry andvariability of the jugular foramen anat-omy (44). The jugular foramen is usuallydivided into a small pars nervosa antero-medially and a larger pars vascularis pos-terolaterally by a dural band or, muchless commonly, a bony septum, which isattached to the jugular spine of the pe-trous bone and jugular process of the oc-cipital bone. Through the pars nervosausually pass the glossopharyngeal nerve(cranial nerve IX), the Jacobson nerve (abranch of cranial nerve IX), and the in-ferior petrosal sinus. Through the parsvascularis usually pass the vagus nerve(cranial nerve X), the Arnold nerve (abranch of cranial nerve X), the spinalaccessory nerve (cranial nerve XI), theinternal jugular vein, the posterior men-ingeal artery, and small meningealbranches of the ascending pharyngeal ar-tery (38–44). Posterior to the occipitalcondyle is the condyloid fossa, an inden-tation on the exocranial surface of theskull base. At its anterior margin, the pos-terior condyloid canal (condylar fora-

Figure 2. Schematics show the anatomic relationships of the occipital condyles with the surrounding neurovascular structures. A, Posterior viewof a coronal section passing through the occipital bone and the posterior arches of the atlas and axis. B, Posterior view of a coronal section passingthrough the occipital condyles and the posterior arches of the atlas and axis anterior to the section in A after dissection of the spinal cord, vertebralartery, dura mater, posterior longitudinal ligament, and tectorial membrane. C, Three-quarter view of sphenoid, temporal, and occipital (anteriorand lateral portions) bones. BA 5 basilar artery, HC 5 hypoglossal canal, JF 5 jugular foramen, JT 5 jugular tubercle, OB 5 occipital bone,OC 5 occipital condyle, SC 5 spinal cord, SS 5 sigmoid sinus, VA 5 vertebral artery, VP 5 venous plexus surrounding the vertebral artery.(Modified and reprinted, with permission, from reference 53.)

Volume 216 z Number 3 Occipital Condylar Fractures: A Review z 637

men) is commonly identified, throughwhich pass anastomotic venous channelsfrom the sigmoid sinus to the suboccipi-tal venous plexus (45).

It is clear that a displaced and migratedfragment resulting from an OCF, as wellas a through-and-through fracture in-volving the hypoglossal canal and/or thejugular foramen, can produce impinge-ment on the medulla oblongata and/orvascular structures and/or lower cranialnerves.

MECHANISMS OF INJURY

The CCJ functions as a unit during flex-ion, extension, lateral bending, and axialrotation; however, it has a limited andspecific range of motion (35,46,47). Theconfiguration of the occipitoatloid jointallows good flexion-extension and somelateral bending but negligible axial rota-tion. At this level, flexion is limited byosseous contact of the anterior portion ofthe foramen magnum with the odontoidprocess; however, the major mechanicalstability is largely dependent on the in-

tegrity of the tectorial membrane, whichlimits extension, and alar ligaments,which limit axial rotation and lateralbending. The configuration of the atlan-toaxial and atlanto-odontoid joints pro-vides good and extensive axial rotation;however, less flexion-extension, and es-sentially no lateral bending, is allowed.At this level, the major mechanical sta-bility is provided through the odontoidprocess and the ring of anatomic struc-tures surrounding it (anterior arch of theatlas anteriorly and laterally, transverseligament of the atlas posteriorly). Flexionis further limited by the tectorial mem-brane, extension is limited by the tecto-rial membrane and other posterior struc-tures, and axial rotation is limited by thealar ligaments (35). The configuration ofthe atlas favors its role as a bearing be-tween the occipital condyles and the su-perior articular facets of the axis, withcervical movements basically determinedby the occipital condyles and the axis(48). The only direct connections be-tween the occiput and the axis are thetectorial membrane, the paired alar liga-

ments, the apical ligament of the dens,and the ascending band of the cruciformligament (Fig 1).

From these considerations, as well asthe anatomic description, it is readily dis-cernible that the stability of the CCJ ismuch more dependent on the integrityof the ligamentous structures than on theremaining structures (35,46,47). In caseof disruption of the ligaments runningfrom the occiput to the atlas, some at-tachments remain through the odontoidligaments. Similarly, in case of failure ofthe ligaments running between the atlasand axis, some attachments of the axis toocciput still remain. Most CCJ unstableinjuries result from the destruction of anumber of ligaments in both occipitoat-loid and atlantoaxial joints. At the oc-cipitoatloid joint, the most importantstructures for mechanical stability are thetectorial membrane and the paired alarligaments. Division of these structures incadavers resulted in destabilization of theoccipitoatloid joints so that dislocationof the skull on the atlas could occur (46).

Another major consideration is that

Figure 3. Schematics show the types of OCF according to the classification system of Anderson and Montesano (8). a, Coronal and b, transverseviews from below show a type I OCF, which is a comminuted fracture of the occipital condyle (black arrows) with minimal or no fragmentdisplacement into the foramen magnum. c, Coronal and d, transverse views from below show a type II OCF, which is a basilar skull fracture(arrowheads) extending into the occipital condyle. e, Coronal and f, transverse views from below show a type III OCF, which is a fracture with afragment displaced medially from the inferomedial aspect of the occipital condyle (white arrows) into the foramen magnum.

638 z Radiology z September 2000 Leone et al

trauma involving the CCJ is also influ-enced by the mass and position of theskull in relation to the long axis of thecervical spine at the time of injury (47).Rarely is a direct axial load supplied tothe spine itself; rather, it is transferredfrom the skull base down through thecervical spine. The location of the forceapplied to the skull determines the forcestransferred to the cervical spine (theseinclude axial loading or asymmetric axialloading with lateral bending, symmetricor asymmetric forces applied to the pos-terior occiput, and hyperflexion or hy-perextension forces, in association withdistraction and lateral rotation forces).However, combined injuries are ex-tremely frequent. For example, in motorvehicle accidents, especially rear-end col-lisions, the head, initially slightly ro-tated, will go into maximal rotation fol-lowed by a “whiplash” movement caused

by the impact. In this particular mecha-nism of injury, the alar ligaments, whichlimit axial rotation, are most vulnerable(38).

Finally, the configuration of the CCJresulting from the normal sagittal andtransverse diameters of the foramen mag-num and cervical spinal canal (the upperportion of which is wider than the lowerportion with a relatively greater space forthe upper spinal cord) explains how trau-matic injuries with fragment displace-ment can occur with fewer neurologicdefects than occur in traumatic lower cer-vical spine injuries (35). Severe craniocer-vical injuries, with or without substantialoccipitoatloid dislocations, even thoughunstable, may occur with no neurologicdamage (48–50). However, this is compli-cated by the fact that massive head injuryand intracranial trauma often accom-pany upper cervical injuries.

CLASSIFICATION SYSTEMS

In 1987, Saternus (7) attempted to clas-sify OCFs on the basis of the forms ofstrain applied. However, the most widelyused classification system is the one pro-posed in 1988 by Anderson and Monte-sano (8) who divided OCFs into threetypes, depending on their morphologyand mechanism of injury (Fig 3). Type I isan impaction-type fracture resulting in acomminution of the occipital condyle,with or without minimal fragment dis-placement (Figs 3a, 3b, 4). The mecha-nism of injury is believed to be axial load-ing of the skull onto the atlas, similar toa Jefferson fracture of the atlas, with orwithout lateral bending. It is considered astable entity because the tectorial mem-brane and contralateral alar ligament areintact; however, bilateral lesions may beunstable (1). A type II OCF is part of amore extensive basioccipital fracture, in-volving one or both occipital condyles(Figs 3c, 3d, 5–7). The mechanism of in-jury is a direct blow to the skull. An intacttectorial membrane and alar ligamentspreserve stability. A type III OCF is anavulsion type of fracture near the alarligament resulting in medial fragmentdisplacement from the inferomedial as-pect of the occipital condyle into the fo-ramen magnum (Figs 3e, 3f, 8–10). Themechanism of injury is forced rotation,usually combined with lateral bending.After occipital condylar avulsion, thecontralateral alar ligament and tectorialmembrane may be stressed and “loaded,”resulting in a partial tear or complete dis-ruption. Thus, the type III OCF is consid-ered a potentially unstable injury. Theinferior portion of the clivus may be dis-rupted too (50).

Recently, Tuli et al (5) proposed a newclassification system for the managementand treatment of OCFs, based on the ab-sence or presence of fragment displace-

Figure 4. Type I OCF in a 39-year-old womaninvolved in a motor vehicle accident as adriver. Glasgow Coma Scale score at admis-sion was 13, without lower cranial nerve palsy.(a) Transverse and (b) direct coronal CT scansdemonstrate an impaction fracture of the in-feromedial aspect of the right occipital condyle(arrow) with minimal fragment displacement.Associated injuries included facial fractures(not shown). The patient was treated with asoft cervical collar.

Figure 5. Type II OCF in an 18-year-old maninvolved in a motor vehicle accident as a pe-destrian. Glasgow Coma Scale score at admis-sion was 8, without lower cranial nerve palsy.(a) Transverse CT scan shows a linear fractureof the left occipital condyle (arrowhead),which is an extension of a comminuted skullbase fracture (arrow). (b) Transverse CT scandemonstrates a bone fragment (arrow) medialto the left jugular foramen. (c) Two-dimen-sional oblique sagittal reformation CT imageclearly demonstrates the craniocaudal extentof the fracture line (arrow), which involves theosseous ring of the hypoglossal canal. Associ-ated injuries included cortical contusions anda wedge fracture of C6 (not shown). The pa-tient was treated with a halo vest.

Volume 216 z Number 3 Occipital Condylar Fractures: A Review z 639

ment and stability of the CCJ as assessedwith radiographic, CT, or magnetic reso-nance (MR) imaging evidence of liga-mentous injury (Fig 11) (5,35,51). Tuli etal divided OCFs into type 1, or undis-placed, and type 2, or displaced. Type 2OCFs are subdivided into type 2a if noligamentous injury is detected and type2b if ligamentous injury is detected.Types 1 and 2a are considered to be stablelesions, whereas type 2b is considered tobe unstable. This functional classificationsystem considers OCFs as part of the widespectrum of craniocervical injuries, with-out considering any distinction in OCFanatomy and morphology. Moreover, itsuggests that avulsion fractures of the oc-cipital condyle and alar ligament injuryrepresent a necessary, but not sufficient,cause of craniocervical instability (52).

CLINICAL PRESENTATION

The clinical presentation of patients withan OCF is highly variable. Most severeneurologic deficits reported in patientswith OCF seem to be related to the sever-ity of head injury rather than to the OCFitself (intraaxial contusion or hematoma,subarachnoid hemorrhage, increased in-tracranial pressure, etc).

Brainstem and vascular lesions are clini-cally rare because they are generally fatal.However, cranial nerve deficits (53,54),hemi- or quadraparesis (1,3,10,23), andsigns and symptoms of vertebrobasilarischemia (Fig 10) should alert both the at-tending physician and the radiologist tothe possibility of an associated OCF.

Lower cranial nerve palsy is the mostfrequently noted neurologic deficit, withvarying combinations ranging from iso-lated paralysis (55–57) to full ninththrough 12th cranial nerves palsies (Col-let-Sicard syndrome) (57–60). However,according to the review by Tuli et al (5),which was mainly based on case reports,only 31% of 51 patients with an OCFhave a lower cranial nerve palsy. Further-more, in the prospective study by Bloomet al (1), only one of the nine patientswith an OCF had a lower cranial nervepalsy. The latter rate of occurrence re-flects our experience. In the review byTuli et al, the manifestations of lowercranial nerve palsy were immediate aftertrauma in 63% of the cases (5,10,58),with a delay of a few days to months in37% (5,55,61). It has been hypothesizedthat a delayed presentation is secondaryto the bone healing process and nervepressure due to callous formation or mo-bilization of a bone fragment that was

not adequately stabilized initially (55). Inassociation with lower cranial nerve pal-sies, other symptoms have occasionallybeen reported, such as dysphagia fromretropharyngeal hematoma (62) or torti-collis from concomitant atlantoaxial ro-tatory fixation (Fig 6) (11). High cervicalpain and impaired skull mobility withoutloss of consciousness or neurologic defi-cits have been described in associationwith OCFs (31,49,50,63,64). Patients aregenerally unconscious; however, somemay remain conscious and responsive(8,49,54).

SCREENING

Because of the high variability in clinicalpresentation, as well as the lack of speci-ficity of signs and symptoms, diagnosticimaging is essential for the diagnosis ofOCFs. Skull and cervical spine radio-graphs obtained routinely in patientswith multiple trauma generally do notshow any abnormality because of facialskeleton superimposition. CT is themethod of choice for the diagnosis ofOCFs. But when should a CT study ofthe CCJ to search for an OCF be per-formed? Some authors (2,3) have sug-gested the following as parameters thatshould arouse suspicion for an OCF: pres-ence of posttraumatic palsies of cranialnerves IX, X, XI, or XII; retropharyngealor prevertebral soft-tissue swelling; occip-

ital skull base fracture; fracture or dislo-cation of the axis or atlas; posttraumaticspasmodic torticollis; or unexplained persis-tent posttraumatic upper-neck pain withnormal conventional radiographs. How-ever, the patient’s consciousness may beso impaired that detailed testing of cra-nial nerves is not possible (33). Further-more, as already stated, cranial nervepalsy should be considered uncommonand may be related to different mecha-nisms altogether (brainstem contusion,reversible ischemia, etc). Additional causesof neck pain and prevertebral soft-tissueswelling are not uncommon in patientswho have experienced severe trauma,such as those associated with midfacialfractures (65). We think, as other authorsdo (1,2,4), that there are no specific pre-dictors of OCF. Clinical findings aregenerally inconsistent, and, in our expe-rience, most patients had mild to moder-ate Glasgow Coma Scale scores. Thus, weagree that, despite the clinical presenta-tion and Glasgow Coma Scale score, OCFmust be suspected in all patients sustain-ing high-energy blunt trauma to the headand/or the upper cervical area resultingfrom axial loading or rotation, lateralflexion or bending, and/or direct blow(1,8).

Figure 6. Type II OCF in a 17-year-old girlinvolved in a motor vehicle accident as a pas-senger. Glasgow Coma Scale score at admis-sion was 5, without lower cranial nerve palsy.(a, b) Transverse CT scans (1-mm collimation)(a) through the occipital condyles and (b) 6mm cranial to a show an extensive commi-nuted right skull base fracture (arrowhead in a)extending into the ipsilateral occipital condyle(arrows in a) and involving the osseous ring ofthe ipsilateral jugular foramen (open arrows inb), with evidence of a distorted fragment (solidarrow in b). (c) Two transverse CT imagesthrough C1 and C2 vertebrae are superim-posed to demonstrate associated mild atlanto-axial rotatory subluxation. Associated injuriesincluded right petrous temporal bone fracturesand pneumocephalus (not shown). The pa-tient was treated with a Philadelphia cervicalcollar.

640 z Radiology z September 2000 Leone et al

RADIOLOGIC DIAGNOSIS

The radiographic evaluation of the CCJhas traditionally been difficult, and anumber of lesions may be undetected orpoorly understood (65,66). Occipitalcondyles may be visualized on some skullviews; however, it is difficult to clearly

define basilar skull fractures on standardradiographs obtained in traumatized pa-tients, such fractures having been re-ported in only 20% of cases (67). Bothanteroposterior and lateral (cross-tabletechnique) views of the cervical spine failto depict the occipital condyles becauseof superimposition of the maxilla andocciput on the anteroposterior view andsuperimposition of the occipital condylesthemselves and, frequently, of mastoidprocesses on the lateral view. Further-more, the evaluation of a prevertebralsoft-tissue shadow on the lateral viewmay be limited by many factors such aspatient positioning, endotracheal intu-bation, or adenoidal tissue. The open-mouth (odontoid) view is designed todemonstrate the atlantoaxial relation-ship in the anteroposterior projectionand may include the occipital condylesand the occipitoatloid joints; however,this view is impossible to obtain in pa-tients who are unconscious or intubatedor have severe mandibular or facial inju-ries (65,68,69). In the past, the diagnosisof OCF was obtained with conventionalor complex motion tomography; how-ever, discrete impaction fractures couldbe missed.

Since its introduction, CT has beenconsidered the method of choice for thediagnosis of OCF; however, a routinebrain CT examination in traumatized pa-tients may not enable detection of thesefractures (61). Rather than brain CT toassess a potential intracranial injury (70),cranial CT including the CCJ should beperformed. The CT study should rou-tinely include 5-mm-thick transverse im-ages beginning at the lower border of C2.Images must be reviewed with windowwidth and level optimized for the evalu-ation of both brain parenchyma and os-seous structures. If required, the scandata can then be reviewed again using ahigh-spatial-resolution, bone- or edge-enhancement reconstruction algorithm.In this way, important information canbe obtained in a very short period oftime. In cases with a high index of suspi-cion, the radiologist must complementthe examination with a study of theCCJ. A thin-section technique is themethod of choice to evaluate the trauma-tized CCJ (71). Generally, 1–2-mm-colli-mation transverse sections, 1-mm tableindexing, 1-second scanning, 200–240mAs, 120 kV, 12- to 14-cm display field ofview, and a high-spatial-resolution algo-rithm are best for assessment of skull baseand CCJ anatomy. Both direct transversescanning and two-dimensional multipla-nar reformations are strongly recom-

Figure 7. Type II OCF in a 23-year-old woman who fell from a horse. Glasgow Coma Scale scoreat admission was 10, without lower cranial nerve palsy. These (a) 5-mm and (b) 1-mm collima-tion transverse CT scans clearly demonstrate a left basilar skull fracture (open arrow in a)extending ipsilaterally through the jugular foramen and into the occipital condyle (solid arrows).Associated injuries included right subdural hematoma and a wedge fracture of T12 (not shown).The patient was treated with a hard cervical collar.

Figure 8. Type III OCF in a 16-year-oldboy involved in a motorcycle accident as apassenger. Glasgow Coma Scale score at ad-mission was 8, without lower cranial nervepalsy. The patient had no associated intra-cranial lesions, spinal fractures, or systemicinjuries. CT scans and MR images of theCCJ were obtained 2 weeks after trauma,after clinical and radiographic examina-tions proved stability of the cervical spine,and during treatment with hard cervicalcollar. (a) Transverse and (b) direct coronalCT scans show an avulsion fracture (arrow)of the inferomedial aspect of the right oc-cipital condyle at the insertion site of theipsilateral alar ligament, with substantialmedial displacement of the bone fragment(compare with Fig 4 in which an impactionfracture resulted in only minimal displace-ment of the bone fragment). (c) CoronalT1-weighted MR image shows a subtle areaof high signal intensity (arrowhead) in thesoft tissues medial to the bone fragment(solid arrow), with retraction of the atlantalportion of the ipsilateral alar ligament. Thisarea showed high signal intensity also onthe T2*-weighted images (not shown) andwas most likely due to edema, inflamma-tion, and hemorrhage at the site of liga-mentous injury. The open arrows indicatethe normal atlantal portion of the left alarligament.

Volume 216 z Number 3 Occipital Condylar Fractures: A Review z 641

mended for the most accurate assessmentof the type of fracture and degree of CCJdisplacement (72). Direct coronal scan-ning is not advisable in unstable patientsor in patients with potential or con-firmed spinal fractures and/or systemicinjuries. Three-dimensional shaded-sur-face reconstruction CT images are dra-matically impressive in displaying thefracture and detailing its extent; how-ever, their diagnostic value is limited(73,74). Helical (spiral) CT scanning en-ables high-quality coronal and sagittalreformations and three-dimensional re-constructions of overlapping transverseimages, with a minimum of motion arti-fact and without additional radiation ex-posure. The study may be obtained witha collimation of 1 mm, a pitch of 1 mm,and a section reconstruction interval of 1mm. It is advisable to obtain the CT scanas soon as possible after craniocervicaltrauma for early detection of an OCF andanticipate possible complications thatmay be clinically silent. Follow-up CT isthen recommended, 10 or 12 weeks afterinjury, to document fracture healing (3).

MR imaging does not yield relevant ad-ditional diagnostic information concern-ing the OCFs but is the best ancillarydiagnostic tool complementing CT forevaluation of associated soft-tissue cranio-cervical trauma. The application of MRimaging in the assessment of ligamen-tous structures, particularly the tectorialmembrane and the transverse ligamentof the atlas, is well established and con-tinually increasing (36,50,51,75). MR im-aging is extremely valuable for the eval-uation of the fractured segment inrelation to the surrounding structures (ie,the cerebrospinal fluid spaces, brainstem,and neurovascular structures). In thosepatients with suspected vascular injury,the use of MR angiography may preventthe necessity of conventional angiogra-phy (Fig 10) (3,10,18,58). Additionally,MR imaging is better than CT for theassessment of associated brain and brain-stem injuries, as well as for intracranial

Figure 9. Type III OCF in a 24-year-oldwoman involved in a motor vehicle accidentas a passenger. Glasgow Coma Scale score atadmission was 7, with cranial nerves IX and Xpalsy. Transverse CT scans (1-mm collimation)(a) through the occipital condyles and (b) 3mm cranial to a show left OCF (arrows in a)with fragment displacement (arrowhead in b)into the foramen magnum. Associated injuriesincluded lateral craniocervical subluxation(not shown). The patient was treated with oc-cipitocervical fusion.

Figure 10. Type III OCF in a 48-year-old man involved in a motor vehicle accident. He wasadmitted to another institution where CT scans suggested a fracture of the left occipital condylewith superomedial displacement of a bone fragment. The patient was treated with a hard cervicalcollar and, 3 months later, was transferred to our institution for further treatment. On admission,the patient was conscious and had a left Collet-Sicard syndrome and signs of cerebellar dysfunc-tion. (a) Transverse CT scan and (b) surface rendered three-dimensional CT reformation fromabove demonstrate a healed left OCF with medial upward displacement of a bone fragment(asterisk). (c) Coronal T1-weighted MR image better depicts the displacement of the left occipitalcondylar fragment (black asterisk) in the posterior fossa, impinging on the medulla (arrowheads);a left cerebellar infarction is clearly seen (white asterisk). (d) Frontal reconstruction of a three-dimensional time-of-flight MR angiogram shows that the distal left vertebral artery is distinctlynarrowed and displaced (small arrows) by the medial upward displacement of the left occipitalcondylar fragment. The source images (not shown) demonstrated the related cranial displace-ment and entrapment of the caudal loop of the left posteroinferior cerebellar artery and lack ofevidence of the left anteroinferior cerebellar artery. Large arrow 5 right anteroinferior cerebellarartery.

642 z Radiology z September 2000 Leone et al

hemorrhage, although CT is still consid-ered the current standard for the evalua-tion of acute subarachnoid hemorrhage(76,77).

MANAGEMENT

To date, management of OCFs has notbeen well established because of thesmall number of cases described in theliterature, as well as the lack of prospec-tive studies investigating follow-up of thedifferent treatment modalities. Further-more, the long-term implications of OCFsare not well known. In the review by Tuliet al (5), four of the six patients who didnot receive treatment developed deficits,such as delayed cranial nerves IX throughXII or IX and X palsy (53) and delayed(55) or fluctuating (19) isolated cranialnerve XII palsy (5). The therapeutic strat-egy is generally conservative. The needfor surgery is controversial and has beenadvocated for craniocervical stabilizationand/or neurovascular decompression.Treatment is initially directed toward re-duction and stabilization with externalfixation. Most authors treat Andersonand Montesano types I and II OCFs witha semirigid or rigid cervical collar andtype III with a rigid cervical collar, halo

traction vest, or surgical fixation (1,3,61).Tuli et al (5) have proposed that undis-placed OCFs do not require immobiliza-tion, displaced OCFs with a stable CCJmay be treated with a hard cervical col-lar, and displaced OCFs with an unstableCCJ require rigid external fixation or sur-gical fixation. However, craniovertebralsubluxation is usually treated with cervi-cal traction and early immobilization in ahalo vest (71). The halo traction vest al-lows adjustments in reduction and helpsmaintain the correct position during andafter surgery. Surgical fixation of the CCJis performed by means of posterior fusion(occipitoatlantoaxial arthrodesis). Thetwo common approaches involve the useof a bone graft with or without wire (78–81), but several technical innovationshave been applied to these two conven-tional methods (82,83).

Removal of a fracture fragment com-pressing the vertebral artery and/or thebrainstem has been performed in pa-tients with a stable CCJ (10,84); however,most authors suggest that conservativetherapy may suffice, even when brain-stem compression is present (3).

CONCLUSION

Accurate determination of the true inci-dence of OCF is difficult because the trau-matized patient may be asymptomaticor the condition may be masked by deathor concomitant injuries or may be de-layed in manifestation. Nevertheless, OCFsshould not be considered uncommon,occurring possibly in as many as 16%of patients with craniocervical injury.OCF should be suspected in all patientssustaining high-energy craniocervicaltrauma from an appropriate mechanismof injury (ie, high-energy blunt trauma tothe head and/or neck involving compo-nents of either axial compression, lateralbending, axial rotation, or direct blow),regardless of the clinical condition andphysical examination results. The greatpotential of these fractures for long-termmorbidity due to pain and limited mo-tion, serious neurologic deficits, or evendeath explains the rising therapeutic andmedicolegal implications of an accuratediagnosis. Clinically, an OCF is generallysuspected in patients immediately show-ing symptoms of lower cranial nervepalsy, because of the lack of specificity ofbrainstem-related symptoms. OCF is lesslikely to be suspected when the neuro-logic deficit is delayed, or even lesserwhen the patient is unconscious or neu-rologically intact. Besides a CT study as-

sessing potential intracranial injuries,these patients also require a CT study ofthe CCJ.

Acknowledgment: We thank Massimo Rollo,MD, for his assistance in providing the sche-matic diagrams presented in this article.

References1. Bloom AI, Neeman Z, Simon Slasky B, et al.

Fracture of the occipital condyles and associ-ated craniocervical ligament injury: inci-dence, CT imaging and implications. ClinRadiol 1997; 52:198–202.

2. Clayman DA, Sykes CH, Vines FS. Occipitalcondyle fractures: clinical presentation andradiologic detection. AJNR Am J Neuroradiol1994; 15:1309–1315.

3. Young WF, Rosenwasser RH, Getch C, Jallo J.Diagnosis and management of occipital con-dyle fractures. Neurosurgery 1994; 34:257–261.

4. Noble ER, Smoker WRK. The forgotten con-dyle: the appearance, morphology, and clas-sification of occipital condyle fractures. AJNRAm J Neuroradiol 1996; 17:507–513.

5. Tuli S, Tator CH, Fehlings MG, Mackay M.Occipital condyle fractures. Neurosurgery1997; 41:368–376.

6. Goldstein SJ, Woodring JH, Young AB. Oc-cipital condyle fracture associated with cer-vical spine injury. Surg Neurol 1982; 17:350–352.

7. Saternus KS. Forms of fractures of the occip-ital condyles. Z Rechtsmed 1987; 99:95–108.

8. Anderson PS, Montesano XP. Morphologyand treatment of occipital condyle fractures.Spine 1988; 13:731–836.

9. Schliack H, Schaefer P. Hypoglossal and ac-cessory nerve paralysis in a fracture of theoccipital condyle. Nervenarzt 1965; 36:362–364.

10. Bozboga M, Unal F, Hepgul K, Izgi N, TurantanI, Turker K. Fracture of the occipital condyle.Spine 1992; 17:1119–1121.

11. Bridgman SA, McNab W. Traumatic occipitalcondyle fracture, multiple cranial nerve pal-sies, and torticollis: a case report and reviewof the literature. Surg Neurol 1992; 38:152–156.

12. Bell C. Surgical observations. MiddlesexHosp J 1817; 4:469–470.

13. Kissinger P. Luxationsfraktur im atlantooc-cipital gelenke. Zentralbl Chir 1900; 37:933–934.

14. Ahlgren P, Mygind T, Wilhejelm B. Eine seltenvorkommende fractura basis cranii. FortschrGeb Roentgenstr Nuklearmed 1962; 97:388–391.

15. Camassa NW, Casavola C, Castelli M, et al.Frattura del condilo occipitale. Radiol Med(Torino) 1983; 63:154–155.

16. Peeters F, Verbeeeten B. Evaluation of occip-ital condyle fracture and atlantic fracture:two uncommon complications of cranio-ver-tebral trauma. Rofo Fortschr Geb RontgenstrNeuen Bildgeb Verfahr 1983; 138:631–633.

17. Spencer JA, Yeakley JW, Kaufman HH. Frac-ture of the occipital condyle. Neurosurgery1984; 15:101–103.

18. Curri D, Cervellini P, Zanusso M, BenedettiA. Isolated fracture of occipital condyle: casereport. J Neurosurg Sci 1988; 32:157–159.

19. Deeb ZL, Rothfus WE, Goldberg AL, DaffnerRH. Occult occipital condyle fractures pre-senting as tumors. J Comput Assist Tomogr1988; 12:261–263.

20. Valaskatzis EP, Hammer AJ. Fracture of theoccipital condyle: a case report. S Afr Med J1990; 77:47–48.

Figure 11. Imaging criteria for CCJ instability(5,35). On the basis of the results of radiogra-phy, CT, and/or MR imaging, CCJ is consid-ered to be stable if none of the criteria aredetected and unstable if a single criterion or acombination of them are detected (5).

Volume 216 z Number 3 Occipital Condylar Fractures: A Review z 643

21. Massaro F, Lanotte M. Fracture of the occip-ital condyle. Injury 1993; 24:419–420.

22. Bettini N, Malaguti MC, Sintini M, Monti C.Fractures of the occipital condyles: report offour cases and review of the literature. Skel-etal Radiol 1993; 22:187–190.

23. Emery E, Saillant G, Ismail M, Fohanno D,Roy-Camille R. Fracture of the occipital con-dyle: case report and review of the literature.Eur Spine J 1995; 4:191–193.

24. Alker GJ, Oh YS, Leslie EV. High cervicalspine and craniocervical junction injuries infatal traffic accidents: a radiological study.Orthop Clin North Am 1978; 9:1003–1010.

25. Bucholz RW, Burkhead WZ. The pathologi-cal anatomy of fatal atlanto-occipital dislo-cations. J Bone Joint Surg Am 1979; 61:248–250.

26. Adams VI. Neck injuries. I. Occipitoatlantaldislocation: a pathologic study of twelve traf-fic fatalities. J Forensic Sci 1992; 37:556–564.

27. Adams VI. Neck injuries. II. Atlantoaxial dis-location: a pathologic study of 14 traffic fa-talities. J Forensic Sci 1992; 37:565–573.

28. Adams VI. Neck injuries. III. Ligamentousinjuries of the craniocervical articulationwithout occipito-atlantal or atlanto-axialfacet dislocation: a pathologic study of 21traffic fatalities. J Forensic Sci 1993; 38:1097–1104.

29. Miltner E, Kallieris D, Schmidt G, Muller M.Verletzungen der schadelbasiscondylen beitodlichen strassenverkehrsunfallen. Z Re-chtsmed 1990; 103:523–528.

30. Link TM, Schureier G, Hufendiek A, et al.Substantial head trauma: value of routine CTexamination of the craniocervicum. Radiol-ogy 1995; 196:741–745.

31. Blacksin MF, Lee HJ. Frequency and signifi-cance of fractures of the upper cervical spinedetected by CT in patients with severe necktrauma. AJR Am J Roentgenol 1995; 165:1201–1204.

32. Warwick R, Williams PL, eds. Gray’s anat-omy. 35th British ed. Philadelphia, Pa: Saun-ders, 1973; 285–288.

33. Lustrin ES, Robertson RL, Tilak S. Normalanatomy of the skull base. NeuroimagingClin N Am 1994; 4:465–478.

34. Samii M, Draf W. Surgery of the skull base:an interdisciplinary approach. Berlin, Ger-many: Springer-Verlag, 1989.

35. White AA III, Panjabi MM. Clinical biome-chanics of the spine. 2nd ed. Philadelphia,Pa: Lippincott, 1990.

36. Ellis JH, Martel W, Lillie JH, Aisen AM. Mag-netic resonance imaging of the normal cra-niovertebral junction. Spine 1991; 16:105–111.

37. Schweitzer ME, Hodler J, Cervilla V, ResnickD. Craniovertebral junction: normal anat-omy with MR correlation. AJR Am J Roent-genol 1992; 158:1087–1090.

38. Saldinger P, Dvorak J, Rahn BA, Perren SM.Histology of the alar and transverse liga-ments. Spine 1990; 15:257–261.

39. Parke WW. The vascular relations of the up-per cervical vertebrae. Orthop Clin NorthAm 1978; 9:879–889.

40. Daniels DL, Williams AL, Haughton VM. Jug-ular foramen: anatomic and computed to-mographic study. AJR Am J Roentgenol1984; 142:153–158.

41. Lo WWM, Solti-Bohman LG. High-resolu-tion CT of the jugular foramen: anatomy andvascular variants and anomalies. Radiology1984; 150:743–747.

42. Weber AL, McKenna MJ. Radiologic evalua-tion of the jugular foramen: anatomy, vascu-lar variants, anomalies, and tumors. Neuro-imaging Clin N Am 1994; 4:579–598.

43. Harnsberger HR. Contemporary imaging is-sues for cranial nerves IX through XII: corecurriculum in neuroradiology, part III—headand neck radiology. Presented at the 31stAnnual Scientific Conference and Postgrad-uate Course of the American Society of Headand Neck Radiology. Toronto, Ontario, Can-ada. May 15–18, 1997.

44. Tekdemir I, Tuccar E, Aslan A, et al. Thejugular foramen: a comparative radioana-tomic study. Surg Neurol 1998; 50:557–562.

45. Weissman JL. Condylar canal vein: unfamil-iar normal structure as seen at CT and MRimaging. Radiology 1994; 190:81–84.

46. Werne S. Studies in spontaneous atlas dislo-cation. Acta Scand Orthop 1957; 23(suppl):1–50.

47. Levine AM, Edwards CC. Traumatic lesionsof the occipitoatlantoaxial complex. Clin Or-thop 1989; 239:54–68.

48. Mann FA, Cohen W. Occipital condyle frac-ture: significance in the assessment of occipi-toatlantal stability. AJR Am J Roentgenol1994; 163:193–194.

49. Stroobants J, Fidlers L, Storms JL, Klaes R,Dua G, Van Hoye M. High cervical pain andimpairment of skull mobility as the onlysymptoms of an occipital condyle fracture:case report. J Neurosurg 1994; 81:137–138.

50. Ide C, Nisolle JF, Misson N, et al. Unusualoccipitoatlantal fracture dissociation with noneurological impairment: case report. J Neu-rosurg 1998; 88:773–776.

51. Dickman CA, Mamourian A, Sonntag VKH,Drayer BP. Magnetic resonance imaging ofthe transverse atlantal ligament for the eval-uation of atlantoaxial instability. J Neuro-surg 1991; 75:221–227.

52. Sonntag VKH, Benzel EC, McCormick PC,Fessler RG. Comments to Tuli S, Tator CH,Fehlings MG, Mackay M: occipital condylefractures. Neurosurgery 1997; 41:376–377.

53. Bolender N, Cromwell LD, Wendling L. Frac-ture of the occipital condyle. AJR Am JRoentgenol 1978; 131:729–731.

54. Desai SS, Coumas JM, Danylevich A, Hayes E,Dunn EJ. Fracture of the occipital condyle:case report and review of the literature.J Trauma 1990; 30:240–241.

55. Orbay T, Aykol S, Seckin Z, Ergun R. Latehypoglossal nerve palsy following fracture ofthe occipital condyle. Surg Neurol 1989; 31:402–404.

56. Castling B, Hicks K. Traumatic isolated uni-lateral hypoglossal nerve palsy: case reportand review of the literature. Br J Oral Maxil-lofac Surg 1995; 33:171–173.

57. Demisch S, Lindner A, Beck R, Zierz S. Theforgotten condyle: delayed hypoglossalnerve palsy caused by fracture of the occipi-tal condyle. Clin Neurol Neurosurg 1998;100:44–45.

58. Hashimoto T, Watanabe O, Takase M,Koniyama J, Kobota M. Collet-Sicard syn-drome after minor head trauma. Neurosur-gery 1988; 23:367–370.

59. Wani MA, Tandon PN, Banerji AK, Bhatia R.Collet-Sicard syndrome resulting from closedhead injury: case report. J Trauma 1991; 31:1437–1439.

60. Sharma BS, Mahajan RK, Bhatia S, KhoslaVK. Collet-Sicard syndrome after closed headinjury. Clin Neurol Neurosurg 1994; 96:197–198.

61. Urculo E, Arrazola M, Arrazola M Jr, Riu I,Moyua A. Delayed glossopharyngeal and va-gus nerve paralysis following occipital con-dyle fracture. J Neurosurg 1996; 84:522–525.

62. Mariani PJ. Occipital condyle fracture present-ing as retropharyngeal hematoma. Ann EmergMed 1990; 9:1447–1449.

63. Olsson R, Kunz R. Fracture of the occipitalcondyle as an incidental finding during CT-evaluation of a maxillary fracture. Acta Ra-diol 1994; 35:90–91.

64. Cottalorda J, Allard D, Dutour N. Fracture ofthe occipital condyle. J Pediatr Orthop B1996; 5:61–63.

65. Harris JH Jr, Mirvis SE. The radiology of acutecervical spine trauma. 3rd ed. Baltimore, Md:Williams & Wilkins, 1996; 475–500.

66. Dietemann JL, Wackeneim A. Radiology oftraumatic lesions of the craniovertebral re-gion. In: Samii M, Brihaye J, eds. Traumatol-ogy of the skull base. Berlin, Germany:Springer-Verlag, 1983.

67. Dolan KD, Jacoby CG. Radiology of skullbase fractures. Crit Rev Diagn Imaging 1979;120:101–152.

68. Heinz BC, Textor J, Hansis M. Diagnosticproblems in fractures of the occipital con-dyles. Unfallchirurg 1997; 100:100–104.[German]

69. Katol MH. Cervical spine trauma: what isnew? Radiol Clin North Am 1997; 35:507–532.

70. Zee CS, Go J. CT of head trauma. Neuroim-aging Clin N Am 1998; 8:525–539.

71. Cornelius RS, Leach JL. Imaging evaluationof cervical spine trauma. Neuroimaging ClinN Am 1995; 5:451–463.

72. Wasserberg J, Bartlett RJV. Occipital condylefractures diagnosed by high-definition CTand coronal reconstructions. Neuroradiol-ogy 1995; 37:370–373.

73. Raila FA, Aitken AT, Vickers GN. Computedtomography and three-dimensional recon-struction in the evaluation of occipital con-dyle fracture. Skeletal Radiol 1993; 22:269–271.

74. Savolaine ER, Ebraheim NA, Jackson WT,Rusin JJ. Three-dimensional computed to-mography in evaluation of occipital condylefracture. J Orthop Trauma 1989; 3:71–75.

75. Dickman CA, Greene KA, Sonntag VKH. In-juries involving the transverse atlantal liga-ment: classification and treatment guidelinesbased upon experience with 39 injuries. Neu-rosurgery 1996; 38:44–50.

76. Murray JG, Gean AD, Evans SJ. Imaging ofacute head injury. Semin Ultrasound CT MR1996; 17:185–205.

77. Fiser SM, Johnson SB, Fortune JB. Resourceutilization in traumatic brain injury: the roleof magnetic resonance imaging. Am Surg1998; 64:1088–1093.

78. Harding-Smith J, MacIntosh PK, Sherbon KJ.Fracture of the occipital condyle: a case re-port and review of the literature. J Bone JointSurg Am 1981; 63:1170–1171.

79. Wiesel SW, Rothman RH. Occipito-atlantal hy-permobility. Spine 1979; 4:187–191.

80. Fielding JW. The status of arthrodesis of thecervical spine. J Bone Joint Surg Am 1988;70:1571–1574.

81. Kalfas IH. Cervical spine stabilization: surgi-cal techniques. Neuroimaging Clin N Am1995; 5:491–505.

82. Ransford AO, Crockard HA, Pozo JL, et al.Craniocervical instability treated by con-toured loop fixation. J Bone Joint Surg Br1986; 68:173–177.

83. Grob D, Dvorak J, Panjabi M, et al. Poste-rior occipito-cervical fusion: a preliminaryreport of a new technique. Spine 1990; 16:S17–S24.

84. Jones DN, Knos AM, Sage MR. Traumaticavulsion fracture of the occipital condylesand clivus with associated unilateral atlanto-occipital distraction. AJNR Am J Neuroradiol1990; 11:1181–1183.

644 z Radiology z September 2000 Leone et al