voL. 59-A, NO. I. JANUARY 977 37

Atlanto-Axial Rotatory Fixation

(FIxED ROTATORY SUBLUXATION OF THE ATLANTO-AXIAL JOINT)*

BY J. WILLIAM FIELDING, M.D.t, NEW YORK, N.Y., AND RICHARD J. HAWKINS, M.D.t,

LONDON, ONTARIO, CANADA

F-root thse Department ()/ Orthopedic Surgery. Si. Lukes Hospital Medical Center. New York City

ABSTRACT: In seventeen cases of irreducible

atlanto-axial rotatory subluxation (here called fixa-

tion), the striking features were the delay in diagnosis

and the persistent clinical and roentgenographic de-

formities. All patients had torticollis and restricted,

often painful neck motion, and seven young patients

with long-standing deformity had flattening on one side

of the face. The diagnosis was suggested by the plainroentgenograms and tomograms and confirmed by

persistence of the deformity as demonstrated by

cineroentgenography . Treatment included skull trac-

tion, followed by atlanto-axial arthrodesis if necessary.

of the thirteen patients treated by atlanto-axial ar-

throdesis, eleven had good results, one had a fair re-

suit, and one had not been followed for long enough to

determine the result. Of the remaining four patients,

one treated conservatively had not been followed forlong enough to evaluate the result, two declined sur-

gery, and one died while in traction as the result of

cord transection produced by further rotation of the

atlas on the axis despite the traction.

Rotatory deformities of the atlanto-axial joint are

usually short-lived and easily correctable. Rarely they per-

sist, causing torticollis which is resistant to treatment.

Such persistent rotation was termed rotary fixation of the

atlanto-axial joint by Wortzman and Dewar in 1968. We

prefer the term atlanto-axial rotatory fixation, since the

fixation of the atlas on the axis may occur with subluxation

or dislocation, or when the relative positions of the atlas

and axis are still within the normal range of rotation.

The object of this report is to emphasize that athanto-

axial rotatory fixation may be caused by a variety of condi-

tions and to make suggestions regarding diagnosis and

management based on our experience with seventeen pa-

tients.

Functional Anatomy

The transverse ligament, the primary stabilizer of the

atlanto-axial complex, prevents excessive anterior shift of

the atlas on the axis. The other supporting structures,

* Read at the Annual Meeting of The American Academy of Or-

thopaedic Surgeons, Las Vegas, Nevada, February 4, 1977.1� 105 East 65th Street. New York, N.Y. 10021.�: 450 Central Avenue, Suite 107, London, Ontario N6B 2E8,

Canada. Dr. Hawkins was sponsored by the McLaughlin Foundation ofCanada.

mainly the paired ahar ligaments, are secondary stabilizers

preventing anterior shift l.5.15.20.23#{149} The alas ligaments also

prevent excessive rotation, the predominant motion be-

tween the atlas and the axis, the right ahar ligament limit-

ing left rotation and vice versa

Coutts noted that with an intact transverse ligament,

the athanto-axial articulation pivots on the eccentrically

placed odontoid and complete bilateral dislocation of the

articular processes can occur at approximately 65 degrees

of rotation, with resultant narrowing of the diameter of an

average-sized canal at the level of the atlas to seven mu-

himeters - a reduction sufficient to damage the cord’.

Coutts also noted that with a deficiency of the transverse

ligament allowing five millimeters of anterior displace-

ment of the atlas on the axis, complete unilateral disloca-

tion can occur at 45 degrees of rotation and narrow the

diameter of the canal at the level of the atlas to twelve mih-

himeters.

The vertebral arteries are located so that they are not

affected by the extremes of normal rotation, even though

each vessel is fixed in the foramen transversarium. How-

ever, they can be severely compromised by excessive rota-

tion, particularly if it is combined with anterior displace-

ment. Brain-stem and cerebellar infarction, and even

death, have been reported as the result of excessive head

rotation which damaged these vessels 16.20

Clinical Material

The seventeen patients with rotatory fixation in this

study included nine males and eight females, seven to

sixty-eight years old (average age, 20.6 years). All came

from the New York area, and eleven of them were treated

by the senior author (J.W.F.) at St. Luke’s Hospital Mcdi-

cal Center in New York City. The other six patients wereprovided by Dr. Barnard Jacobs, Dr. Leon Root, and Dr.

p. D. Wilson, Jr. , of The Hospital for Special Surgery in

New York City, and by Dr. G. Dean MacEwen of the

duPont Institute in Wilmington, Delaware. We re-

examined fourteen of the patients personally and the

follow-up information on the other three was obtained

from the charts of The Hospital for Special Surgery.

Clinical Findings

The cases in this series were analyzed with a view to

determin ing the clinical and roentgenographic characteri s-

tics of this rare condition, and to establishing principles

for treatment based on the results obtained.

Anteroposterior and lateral roentgenograms showing Gallie fusion used in this series.

38 J. W. FIELDING AND R. J. HAWKINS

THE JOURNAL OF BONE AND JOINT SURGERY

. : �

�_�____4�.�

FIG. I

Typical cock robin position of rotatory fixation illustrating tilt (lateral

flexion) to one side, rotation toward the opposite side, and slight flexion.

Course

The onset of the deformity was spontaneous in four

patients and was associated with an upper respiratory-tract

infection in five, minor trauma in three, and major trauma

in two.

In the other three patients, onset followed the apphica-

tion of an orthodontic device in one, surgical repair of a

cleft palate in another, and removal of a body cast during

treatment of scohiosis in a patient with neurofibromatosis.The delay in diagnosis in these seventeen patients

ranged from none to twenty-eight months, the average

being 1 1 .6 months. In only two patients was the hesion ac-

curately diagnosed at onset, while in the others a multitude

of diagnoses were made and many treatments were at-

tempted before the cause was correctly identified as a de-

formity of the atlanto-axial joint complex.

Signs and Symptoms

All patients had torticolhis and a diminished range of

motion, and seven had facial flattening. In ten, mild pain

was produced when any attempt was made to correct the

deformity. Using their own neck muscles, all patients

could increase the clinical deformity but could correct it

only to the neutral position or to just beyond neutral. Neck

extension was diminished by approximately 50 per cent.

The typical head position was 20 degrees of tilt to one

side, 20 degrees of rotation to the opposite side, and slight

flexion. This position has been likened to that of a robin

listening for a worm, the so-called cock robin position

(Fig. 1). One patient had weakness of the lower cx-

tremities, up-going toes, and a radicuhopathy of the second

cervical-nerve root. In three patients, the sternocheidomas-

toid muscle on the side from which the head was tilted was

in some degree of spasm, as if attempting to correct the

deformity.

Treatment

Thirteen patients had some form of arthrodesis: dcv-

en, an atlanto-axiah Galhie fusion (Fig. 2); one, an

bccipito-axial arthrodesis because of associated fracture;

and one, a fusion from the occiput to the third cervical yen-

tebra because of widespread bone destruction caused by

rheumatoid arthritis. All patients had preoperative skull

traction for an average of fifteen days, usually in the range

of 4.5 to 6.8 kilograms, in an attempt to correct the defor-

mity.

Clinical reduction was achieved but the amount of

correction roentgenographicahhy was difficult to assess.

Postoperative traction was continued for six weeks, to

maintain as much correction as possible while the fusion

occurred. There were no significant operative or post-operative complications. Three patients were treated cx-

pectantly because two declined surgery and one had only

mild disability.

Follow-up

The patients were followed for from three months to

twelve years (average, 4.2 years). Of the thirteen patients

who had fusion, eleven were asymptomatic and had a

normal head position, no facial flattening, and a functional

range of motion; one, though much improved from her

FIG. 3-A

ATLANTO-AXIAL ROTATORY FIXATION 39

VOL. 59-A, NO. I , JANUARY 1977

preoperative status, had a mild torticollis deformity , mm-

imum facial flattening, and slight discomfort during activ-

ity; and the remaining patient, who had neurofibromatosis,

had not been followed for long enough to evaluate the re-

sult at the time of writing. Loss of rotation was not a sig-nificant complaint, the maximum loss being 25 degrees in

either direction. In all patients, the fusion was clinically

and roentgenographicalhy solid in less than three months.

The two patients who refused surgery were followed for

eight years after the diagnosis was made. One was

asymptomatic with a functional range of motion and no

clinical deformity, even though the roentgenographic de-

formity persisted, and the other patient remained un-

changed from the time of initial diagnosis, retaining both

the cock robin position and the facial flattening. The third

patient, whose deformity was considered too mild to war-

rant treatment, was normal at follow-up.

Another patient in this series was a sixty-five-year-

old woman who twisted her neck while yawning and had

immediate sharp pain and a persistent torticohhis.

Roentgenograms revealed a rotatory deformity and

marked anterior displacement of the atlas on the axis.

After treatment in halter traction with approximately 4.5

kilograms of weight for ten days without reduction, she

died suddenly when she turned her head in the direction of

the rotatory deformity. Autopsy revealed that the atlas had

rotated across the canal, crushing the cord. Although her

deformity was of short duration, her case is included as an

example of rotatory fixation because of the resistance of

the subluxation to correction while in traction. It also illus-trates how these patients can occasionally increase their

rotatory displacement even though they cannot correct it.

Roentgenographic Findings

The roentgenographic features of rotatory fixation

may be confusing because of difficulty in positioning the

patient and interpreting the roentgenograms. Even the

normal upper part of the cervical spine may show consid-

enable variation, due to slight malalignment of the head or

the x-ray beam and the many congenital and devel-

opmental anomalies that occur in this region 3,4.14,18.23#{149}

Based on studies of routine roentgenograms, tomo-

grams, and cineroentgenography, we identified the follow-

ing roentgenographic manifestations of rotation of the

atlas on the axis which are present in any patient with ton-

ticollis.1 . Open-mouth anteropostenior projection.

a. The lateral mass of the atlas that is rotated for-

ward appears wider and chosen to the midline (medial

offset) while the opposite mass appears narrower and

farther away from the midline (lateral offset).

b. On the side where the atlas has rotated back-

wand (right side on right rotation and vice versa), the joint

between the lateral masses of the atlas and axis is some-

times obscured due to apparent overlapping (Figs. 3-A and

3-B).c. In most normal individuals, the spinous pro-

Op en-mouth roentgenogram showing how the lateral mass of the atlasthat has moved forward appears wide and closer to the midline (medialoffset) while the opposite mass appears narrower and away from the mid-line (lateral offset). Note also the poorly outlined articular process on theside that has rotated backward.

FIG. 3-B

Diagram of the atlanto-axial joint viewed from above, showing the re-lationship of the structures seen on the anteroposterior view in the neutralposition (left) and with the atlas rotated to the right (right). (Reprinted bypermission from: Rotary Fixation of the Atlantoaxial Joint: RotationalAtlantoaxial Subluxation, by 0. Wortzman and F. P. Dewar. Radiology,90: 479-487, 1968.)

cess of the axis is not significantly deviated from the mid-

line until rotation of more than 50 per cent of total normal

notation has occurred (that is, deviation to the left with

right rotation and vice versa). However, if any lateral flex-

ion (tilt) is associated with notation of the cervical spine

below the atlas, the spinous process of the axis may appear

to be markedly deviated from the midline; that is, deviated

to the right with left tilt or vice � . Therefore, if the

spinous process of the axis, the best indicator of axial rota-

tion, is tilting in one direction and notated in the opposite

direction, rotatory fixation on torticolhis, whatever the

cause, is present, and usually the chin and the spinous pro-

cess are on the same side of the midline (Fig. 4).

2. Lateral projection.

a. If one wedge-shaped lateral mass of the atlas

40 J. W. FIELDING AND R. J. HAWKINS

THE JOURNAL OF BONE AND JOINT SURGERY

Anteroposterior roentgenogram showing the chin (symphysis menti)

and bifid spine of the axis on the same side of the midline. This relation-ship is due to lateral flexion (rather than simple rotation) which causesconcomitant rotation of the axis, thus deviating the spine of the axis.

FIG. 5

Lateral roentgenogram of the cervical spine showing the wedgedshape of the lateral mass of the atlas that has rotated forward to occupythe position normally held by the oval anterior arch of the atlas. Thisprojection suggests assimilation of the atlas into the occiput. The twohalves of the posterior arch. though not well visualized. are not superim-posed on each other because the head is tilted.

has rotated anteriorly to where the oval anterior arch of the

atlas normally lies, measurement of the atlas-dens interval

may occasionally be difficult but lateral tomograms usu-

Anteroposterior tomogram of the atlanto-axial region erroneouslysuggests that one lateral atlantal mass is absent, but actually it has ro-tated to another plane.

ally resolve the problem. Because anterior displacement of

the atlas may significantly constrict the spinal canal, it is

important to obtain this measurement (Fig. 5).

b. Because of the tilt of the atlas, the two halves

of its posterior arch are not superimposed on each other on

the roentgenogram, which may even suggest assimilation

of the atlas into the skull if the occiput is superimposed on

the tilted posterior arch of the atlas (Fig. 5).

3. Tomograms in the anteroposterior projection may

show the two lateral masses of the atlas to be in different

coronal planes and may suggest erroneously that one hat-

enal mass is absent (Fig. 6).

These roentgenographic manifestations were present

in the patients in this series but they were not diagnostic of

rotatory fixation, only indicating a rotated position of the

atlas with respect to the axis. Their presence should lead to

further investigation.

in our experience, the most useful procedure to dem-

onstnate athanto-axial rotatory fixation is cineroentgenog-

raphy in the lateral projection. This procedure demon-

strates that the posterior arches of the atlas and axis move

as a unit during attempted neck rotation. Normally the

atlas clearly rotates independently on the relatively im-

mobile axis. Of the eight patients in this series who had

this procedure, all had enough motion in the neck to per-

mit demonstration of the fixation of the atlas on the axis

during attempted rotation.Wortzman and Dewar suggested that a persistent

asymmetrical relationship of the dens to the articular mas-ses of the atlas not correctable by rotation is the basic

diagnostic criterion for this condition. This asymmetry can

be demonstrated by obtaining open-mouth roentgeno-

grams with the neck in 15 degrees of rotation to the right

FIG. 7

A young patient with a markedly increased atlas-dens interval and acompensatory severe swan-neck deformity of the lower cervical seg-ments.

ATLANTO-AXIAL ROTATORY FIXATION 41

VOL. 59-A, NO. 1, JANUARY 1977

and to the heft. Although this method of demonstrating

fixation is accurate, it is more easily shown by cinenoent-

genography. If cineroentgenography is not available, the

method of Wortzman and Dewar may be used, but we

have found these noentgenograms difficult to make and

interpret.

If the rotatory fixation is complicated by rupture or

deficiency of the transverse ligament, the atlas-dens inter-

val when the neck is flexed may be greater than three mil-

limetens in older children and adults or greater than four

millimeters in younger children 8,11,13,17 Occasionally, in

the presence of such pathological anterior displacement of

the atlas on the axis, there may be a compensatory swan-

neck deformity of the lower part of the cervical spine

(Fig. 7).

Discussion

In these seventeen patients with this rarely seen de-

formity, the displacement was difficult or impossible to

correct, and in most of them the deformity recurred when

conservative treatment was concluded. As previously

noted, the average delay between onset and diagnosis was

1 1 .6 months. In most of these cases it was not possible to

determine whether earlier diagnosis and aggressive con-

senvative treatment would have prevented this severe form

of rotatory fixation. Certainly delay in diagnosis was not a

factor in the woman who died while in traction.

Although simple support on even observation is usu-

ally all that is needed for atlanto-axial rotatory displace-

ment, the cases in which these simple measures will not

prevent the development of fixed deformity cannot be dif-

ferentiated from those with the common, easily resolvable

rotatory displacement which must be diagnosed primarily

by history and clinical examination, because in the early

stages a satisfactory cineroentgenographic examination is

usually not possible due to pain and muscle spasm.

The importance of recognizing atlanto-axial rotatory

fixation lies in the fact that it may indicate a compromised

atlanto-axial complex with the potential to cause neural

damage or even death.

Long-standing rotatory fixation, like hong-standing

tonticollis for any reason, may cause facial asymmetry in

younger patients. Of the seven patients who had facial

asymmetry in the present series, only one was left with

any stigma at follow-up after fusion. Of the two adults

who had facial flattening when the lesion was diagnosed

during their teen-age years and treated conservatively,

only one had persistent facial flattening seven years after

diagnosis.

Neural involvement, more commonly seen when both

rotatory and anteropostenior displacement were present,

ranged from mild nerve-root irritation causing paresthesias

to gross motor involvement and even fatal cord compres-

sion, as in the case of the sixty-five-year-old woman

already described who died while in traction 9.10.12.21 In

another case in this series, that of a sixty-eight-year-old

woman with rheumatoid arthritis, there was marked poste-

non displacement as well as rotatory fixation of the atlas.

She had paresthesias in the distribution of the greater oc-

cipital nerve and pyramidal-tract signs.

Classification

Based on the seventeen patients in this series, we

classified rotatory fixation into four types. An illustrative

case in each classification is presented to demonstrate the

roentgenographic characteristics of each type and the

many varied features observed in patients with atlanto-

axial rotatory fixation (Fig. 8).

Type I - Rotatory Fixation without Anterior Displacement

ofthe Atlas (Displacement of Three Millimeters or Less)

This was the most common deformity , occurring in

eight patients whose fixed rotation was within the normal

range of atlanto-axial rotation and whose transverse higa-

ment was intact, so that the dens acted as the pivot.

Case Report

A nine-year-old Oriental boy awoke one morning, after swimming

the previous day, with a stiff neck and his head cocked to one side. The

diagnosis of rotatory fixation was not made for eight months, during

which time he was treated with traction both at home and in the hospital,

but the deformity persisted. Examination eight months after onset

showed a typical torticollis and a diminished range of neck motion with

pain at the extremes of motion. Plain roentgenograms and tomography

demonstrated the atlas to be rotated on the axis and cineroentgenography

confirmed the fixation. There was no anterior displacement of the atlas

on the axis demonstrated on flexion-extension roentgenograms. Treat-

ment included Vinke-tong traction with 4.5 kilograms of weight for

three weeks, which partially corrected the deformity, followed by

atlanto-axial fusion. On follow-up after ten years he appeared normal

and was asymptomatic, although his range of neck rotation was de-

42 J. W. FIELDING AND R. J. HAWKINS

THE JOURNAL OF BONE AND JOINT SURGERY

Drawings showing the four types of rotatory fixation. (a) Type I - rotatory fixation with no anterior displacement and the odontoid acting as thepivot; (b) Type II - rotatory fixation with anterior displacement of three to five millimeters, one lateral articular process acting as the pivot; (c) TypeIII - rotatory fixation with anterior displacement of more than five millimeters; and (d) Type IV - rotatory fixation with posterior displacement.

creased approximately 25 degrees in both directions. Roentgenograms

showed a solid fusion.

Type ii - Rotatory Fixation with Anterior Displacement

of the Atlas of Three to Five Millimeters

This was the second most common lesion (five pa-

tients). It was associated with deficiency of the transverseligament and unilateral anterior displacement of one lat-

enah mass of the atlas while the opposite, intact joint acted

as the pivot. In these patients there was abnormal anterior

displacement of the atlas on the axis and the amount of

fixed rotation was in excess of normal maximum rotation.

Case Report

A twelve-year-old boy had his coat pulled over his head during a

fight. The next day he awoke with a typical torticollis and stiff neck.

Treatment included cervical collars and traction in a hospital with head-

halter traction and 4.5 kilograms of weight. Section of the sterno-

cleidomastoid muscle was also recommended. Seven months after onset,

when the diagnosis of atlanto-axial rotatory fixation was made, examina-

tion revealed torticollis, moderate facial flattening, and a diminished

range of motion of the neck. Plain roentgenograms and tomography

demonstrated that the atlas was rotated on the axis, and

cineroentgenography confirmed that the rotation was fixed. Flexion-

extension roentgenograms showed that the atlas was anteriorly displaced

five millimeters on the axis. Treatment included traction in Vinke tongs

with 5.4 kilograms of weight for three weeks, which partially reduced

the deformity to the neutral position, and then atlanto-axial arthrodesis.

At follow-up seven years later, the boy appeared normal and had no

symptoms. His range of neck motion was full and the fusion was solid.

Type III - Rotatory Fixation with Anterior Displacement

of More than Five Millimeters

This type, seen in three patients, was associated with

a deficiency of both the transverse and secondary higa-

ments. Both lateral masses of the atlas were displaced an-

tenionly, one more than the other, producing the notated

position.

Case Report

A thirteen-year-old mongoloid boy was noted to have torticollis fol-

lowing an upper respiratory infection. His previous history was sig-

nificant in that three years earlier he hnd been in a motor vehicle accident

which caused a head injury. Diagnosis was delayed for three months

after onset of the torticollis, during which time he was treated intermit-

tently with skull traction, but the deformity had recurred following re-

moval of the traction. At that time, while his neck was being massaged,

he had pain in the neck. headache, and involuntary evacuation of the

bladder. In addition, on two other occasions he had become cyanotic.

Examination revealed torticollis and markedly diminished neck motion,

with spasm of the sternocleidomastoid muscle on the side from which the

head was tilted. Roentgenograms revealed significant rotation of the

atlas on the axis. Because of the symptoms of neural involvement,

cineroentgenography was not performed. Flexion-extension roentgeno-

grams revealed twelve millimeters of anterior displacement of the atlas

on the axis. Skull traction using Vinke tongs with 4.5 kilograms ofweight for five days was followed by atlanto-axial arthrodesis. At

follow-up ten years later, his appearance was normal and he had a full

range of neck motion. At that time he was participating in athletics.

Type I I’ - Rotatory Fixation with Posterior Displacement

This rare lesion (one patient) occurred when a de-

ficient dens allowed posterior shift of one on both lateral

masses of the atlas, one of them shifting more than the

other so that the atlas was rotated on the axis.

Case Report

A sixty-eight-year-old woman with rheumatoid arthritis and known

anteroposterior instability of the atlanto-axial joint suddenly had torticol-

lis associated with weakness of the lower extremities and numbness in

the distribution of the second cervical nerve. Plain roentgenograms and

tomography demonstrated rotation of the atlas on the axis, an absent

dens, and fifteen millimeters of posterior displacement of the atlas on the

axis. Initially the patient was maintained for three months in halo trac-

tion, which partially reduced the deformity. Because of the neural in-

volvement, fusion from the occiput to the third cervical vertebra was per-

formed. Extension down to the third cervical segment was necessary be-

cause there was so much bone destruction. When the patient was last

seen, three months after operation, the fusion appeared to be solid, but it

was too early for complete evaluation.

This classification provides some guidance as to

prognosis and treatment. Type I is the most benign be-

cause the transverse ligament is intact, and patients with

this lesion can be treated more on less expectantly. Type

II, with a deficient transverse ligament, is potentially

dangerous. Types III and IV are extremely rare but have

catastrophic potential. A Type-Ill fixation caused the only

fatality in the present series, while there was a Type-IV

lesion in the patient with pyramidal-tract involvement. It

is therefore extremely important to recognize Types III

and IV promptly and to initiate proper treatment.

Etiology and Mechanism of Fixation

The cause of the fixation remains obscure, since no

ATLANTO-AXIAL ROTATORY FIXATION 43

VOL. 59-A, NO. I , JANUARY 1977

anatomical or autopsy evidence bearing on this question is

available. Wittek suggested that there is an effusion in the

synoviai joints producing stretching of the ligaments.

Coutts proposed the theory that synovial fringes, when in-

flamed or adherent, may block reduction, while Fiorrani-

Gallotta and Luzzatti believed that the deformity is due to

rupture of one or both alar ligaments and the transverse

ligament. Watson Jones postulated hyperemic decalcifica-

tion with loosening of the ligaments. Grisel suggested that

muscle contracture might follow an upper respiratory-tract

infection and be a factor. Hess and associates concluded

that there is a combination of factors, including muscle

spasm, which prevent reduction in the early stages.

Wortzman and Dewar postulated damage to the atlanto-

axial articular processes of unknown nature. It is our belief

that reduction is probably obstructed in the early stages by

swollen capsular and synovial tissues and by associated

muscle spasm.

If the abnormal position persists because of a failure

to achieve reduction, ligament and capsular contractures

develop and cause fixation.

Rotatory fixation may be associated with anterior or,

rarely, posterior displacement of the atlas on the axis.

With anterior displacement there must be ligament de-

ficiencies, caused by either trauma or infection. The com-

bination of rotatory and anterior displacement is easily cx-

plained but there is no explanation for the obstruction to

reduction.

After fractures with rotatory fixation (as in two pa-

tients in this series), the mechanism of fixation may be

secondary to obvious articular damage, as exemplified by

the following cases.

Case Reports

A fifteen-year-old girl fell while ski-jumping and sustained a frac-

ture of the superior articular process of the axis which caused a refractory

torticollis not corrected by skull traction or a halo cast used intermit-

tently for one year. The atlas was displaced six millimeters anteriorly on

the axis and it was thought that the fixation was related to the ligament

damage and fracture of the articular process.

A forty-year-old man who caught his head in a printing press had a

compression fracture of the left lateral mass of the atlas which was dis-

placed five millimeters anteriorly. Fixation in this patient was also

thought to be related to ligament and bone damage.

Diagnosis

To make the diagnosis of atlanto-axial rotatory fixa-

tion, one must be alert to the possibility of fixation and ap-

preciate the anatomical and roentgenographic charactenis-

tics of the athanto-axial joint. The diagnostic criteria are a

resistant torticolhis and fixation of the atlas on the axis

demonstrated roentgenognaphically.

Unexplained persistent tonticohlis, particularly in

younger patients, should not be dismissed lightly. The dif-

ferential diagnosis includes congenital torticolhis, infec-

tion of the cervical spine, cervical adenitis, congenital

anomalies of the dens, syningomyehia, cerebellar tumors,

bulbar palsies, ocular problems, and so on. An important

finding differentiating spasmodic tonticolhis or so-called

wry neck from this condition is that the shortened sterno-

cleidomastoid muscle is the deforming force and is in

spasm in spasmodic torticohhis, while in rotatory fixation

the elongated stennocleidomastoid may be in spasm (espe-

cially in the early stages) as if attempting to correct the de-

formity.

The usual picture is that of a persistent torticohhis

which began spontaneously after trivial trauma or after an

upper respiratory-tract infection. The diagnosis of fixationin fifteen of the patients in this series was delayed and was

made after many types of treatment had been unsuc-

cessful. Most of these patients were young. A typical case

that emphasizes the diagnostic difficulties is that of a

seven-year-old girl who began to have tonticolhis two

weeks following an ear infection. Traction, physio-

therapy, a Minervajacket, neck manipulation, a halo cast,

and finally a Milwaukee brace had failed to correct the de-

formity. She had been seen by many doctors, including a

psychiatrist, and all the while she had an unrecognized

atlanto-axial rotatory fi xation . C ineroentgenography

twenty-five months after onset confirmed the diagnosis,

and after partial reduction by skull traction atlanto-axial

arthrodesis was performed because of the lesion’s nesis-

tance to correction.

The roentgenographic findings can be confusing.

Plain roentgenognams and tomograms are helpful, since

they often indicate that the atlas is rotated on the axis, but

these findings are not pathognomonic of fixation because

the same relationship may be seen in individuals who are

normal or have torticollis due to other causes. If the plain

roentgenograms or the tomognams show evidence of

atlanto-axiah rotation, further investigation is warranted.

Cineroentgenography will confirm the presence of fixa-

tion. However, open-mouth projections with the neck in

neutral position and in 15 degrees of rotation to the right

and left of the mid-sagittal plane may show the fixation if

cineroentgenography is not available. Flexion-extension

stress noentgenograms will rule out any antenopostenior

displacement.

Treatment

If there is rotatory fixation, atlanto-axial stability may

be compromised and even a minor injury to the neck may

be catastrophic, especially when anterior displacement is

present. Skeletal skull traction of some form (we prefer

Vinke tongs) should be applied. The weight used for skull

traction is dependent generally on the age of the patient:

3.2 or 3.6 kilograms in younger children and up to 6.8

kilograms in adults. This may be increased by increments

of 0.5 to 0.9 kilogram every three or four days if conrec-

tion is not obtained, up to an arbitrary maximum of 6.8

kilograms in children and 9. 1 kilograms in adults. If the

deformity is corrected, the reduction is maintained in some

form of immobilization, such as continued traction or a

Minerva jacket, which should be maintained for three

months. However, since recurrence of the deformity after

44 J. W. FIELDING AND R. J. HAWKINS

ThE JOURNAL OF BONE AND JOINT SURGERY

this interval in our experience is common, patients with

hong-standing fixation (more than three months) are prob-

ably best treated by fusion. Manipulation of these fixed de-

fonmities is not recommended because of the inherent dan-

gers.We believe that fusion should be performed to ensure

stability and to maintain correction when there is neural

involvement (even if transient) on anterior displacement

(particularly if it is more than four millimeters) and when

adequate conservative management has failed to achieve

and maintain connection. Although not used in the present

series, a halo cast might occasionally be helpful.

If fusion is contemplated, traction should be insti-

tuted for two to three weeks preoperativehy to achieve as

much correction of the deformity as possible. Usually the

clinical deformity can be corrected by bringing the head to

neutral . However, persistent roentgenographic abnor-

mahities may be difficult to see. An attempt should be

made to obtain whatever correction is possible, and after

fusion the patient should remain in traction with approxi-

matehy 4.5 to 6.8 kilograms depending on age for six

weeks, to maintain as much connection as possible while

the fusion is becoming solid. Perhaps in this instance as

well, a halo cast would be helpful to permit early ambula-

tion when feasible.

Arthrodesis of the atlanto-axial joint limits rotation,

but seldom by more than 25 degrees in either direction.

Loss of notation was not a significant problem in our pa-

tients, most of whom were young, at an age when com-

pensatory motion develops in the lower part of the cervical

spine. The average loss of notation at follow-up in the pa-

tients in whom this was measured was 1 5 degrees in either

direction.

Conclusions

Persistent tonticohhis in younger patients, particularly

after trivial trauma or an upper respiratory-tract infection,

suggests a diagnosis of atlanto-axial rotatory fixation. The

diagnosis can be confirmed by cineroentgenography. An-

tenon displacement of the atlas, indicating a deficient

transverse ligament, should be ruled out by flexion-

extension lateral roentgenograms. If conservative man-

agement fails to achieve reduction or is followed by a ne-

currence of the deformity, anthrodesis is indicated.

References

1. COUTTS, M. B.: Atlanto-Epistropheal Subluxations. Arch. Surg., 29: 297-311, 1934.2. FIELDING, J. W.: Cineroentgenography of the Normal Cervical Spine. J. Bone and Joint Surg. , 39-A: 1280-1288, Dec. 1957.3. FIELDING, J. W.: Normal and Selected Abnormal Motion of the Cervical Spine from the Second Cervical Vertebra to the Seventh Cervical

Vertebra Based on Cineroentgenography. J. Bone and Joint Surg., 46-A: 1779-1781, Dec. 1964.4. FIELDING, J. W.: Selected Observations on the Cervical Spine in the Child. Curr. Pract. Orthop. Surg., 5: 31-55, 1973.5. FIELDING, J. W.; COCHRAN, G. V. B.; LAWSING, J. F. III; and HOHLJ MASON: Tears ofthe Transverse Ligament ofthe Atlas. A Clinical and

Biomechanical Study. J. Bone and Joint Surg., 56-A: 1683-1691, Dec. 1974.6. FIORRANI-GALLOTTA, GIOVANNI, and LUZZATTI, GUID0: Sublussazione laterale e sublussazione rotatorie dell’atlante. Arch. ortop. . 70: 467-

484, 1957.7. GRISEL, P.: Enucl#{233}ation de l’atlas et torticollis naso-pharyngien. Presse med. , 38: 50-53, 1930.8. GROGONO, B. J. S.: Injuries of the Atlas and Axis. J. Bone and Joint Surg., 36-B: 397-410, Aug. 1954.9. HESS. J. H.: BRONSTEIN, I. P.; and ABELSON, S. M.: Atlanto-Axial Dislocations. Unassociated with Trauma and Secondary to Inflammatory

Foci in the Neck. Am. J. Dis. Child. , 49: 1 137-1 147, 1935.10. HUNTER, G. A.: Non-Traumatic Displacement ofthe Atlanto-Axial Joint. A Report ofSeven Cases. J. Bone and Joint Surg., 50-B: 44-51, Feb.

1968.11. JACKSON, HA�uuS: The Diagnosis of Minimal Atlanto-Axial Subluxation. British J. Radiol., 23: 672-674, 1950.12. JACOBSON, GEORGE, and ADLER, D. C.: Examination of the Atlanto-Axial Joint Following Injury. With Particular Emphasis on Rotational Sub-

luxation. Am. J. Roentgenol. , 76: 1081-1094, 1956.13. MARTEL, WILLIAM: The Occipito-Atlanto-Axial Joints in Rheumatoid Arthritis and Ankylosing Spondylitis. Am. J. Roentgenol. , 86: 223-240,

1961.14. PAUL, W., and M0IR, W. W.: Non-Pathologic Variations in Relationship of the Upper Cervical Vertebrae. Am. J. Roentgenol., 62: 5 19-524,

1949.15. ROY-CAMILLE, R.; DE LA CAFFINIERE, J.-Y.; and SAILLANT, G.: Traumatismes du rochis cervical sup#{235}rieurCl-C2. Paris, Masson, 1973.16. SCHNEIDER, R. C., and SCHEMM, G. W.: Vertebral Artery Insufficiency in Acute and Chronic Spinal Trauma. With Special Reference to the

Syndrome of Acute Central Cervical Spinal Cord Injury. J. Neurosurg. , 18: 348-360, 1961.17. STEEL, H. H.: Anatomical and Mechanical Considerations of the Atlanto-Axial Articulations. In Proceedings of The American Orthopaedic

Association. J. Bone and Joint Surg., 50-A: 1481-1482, Oct. 1968.18. VON TORKLUS, DETLEF, and GEHLE, WALTER: The Upper Cervical Spine. New York, Grune and Stratton, 1971.19. WATSON JONES, R.: Spontaneous Hyperaemic Dislocation of the Atlas. Proc. Roy. Soc. Med. , 25: 586-590, 1932.20. WERNE, SVEN: Studies on Spontaneous Atlas Dislocation. Acta Orthop. Scandinavica, Supplementum 23, 1957.21. WILSON, M. J.; MICHELE. A. A.; and JACOBSON, E.: Spontaneous Dislocation of the Atlanto-Axial Articulation, Including a Report of a Case

with Quadriplegia. J. Bone and Joint Surg. , 22: 698-707. July 1940.22. WITTEK, ARNOLD: Em Fall von Distensionsluxation im Atlanto-epistropheal-Gelenke. Muenchener med. Wochenschr. . 55: 1836-1837, 1908.23. WORTZMAN, G., and DEWAR, F. P.: Rotary Fixation of the Atlantoaxial Joint: Rotational Atlantoaxial Subluxation. Radiology, 90: 479-487,

1968.

本文献由“学霸图书馆-文献云下载”收集自网络，仅供学习交流使用。

学霸图书馆（www.xuebalib.com）是一个“整合众多图书馆数据库资源，

提供一站式文献检索和下载服务”的24 小时在线不限IP

图书馆。

图书馆致力于便利、促进学习与科研，提供最强文献下载服务。

图书馆导航：

图书馆首页 文献云下载 图书馆入口 外文数据库大全 疑难文献辅助工具