Journal of Orthopaedic Research 9828-834 Raven Press, Ltd., New York 0 1991 Orthopaedic Research Society

In Vivo Flexion/Extension of the Normal Cervical Spine J. Dvorak, "M. M. Panjabi, *J. E. Novotny, and "J. A. Antinnes Department of Neurology, Spine Unit of the Schulthess Hospital, Zurich, Switzerland, and *Department of Orthopaedics and Rehabilitation, Yale University School of Medicine, New Haven, Connecticut, U.S.A.

~~~

~~~

~

Summary: Twenty-two women (age range 25-49 years, average 30.9 years) and twenty-two men (age range 23-42 years, average 31.6 years), all healthy and asymptomatic, underwent passive flexion/extension examinations of the cervical spine. Functional x-rays were taken and analyzed using a computerassisted method that quantified intervertebral rotations, translations, and locations of the centers of rotation for each level Cl-C2-C6-C7. The aim of the study was to establish values for these parameters for a normal population as related to age and gender. In the process, a statistically significant difference was found in the average value of rotation between male and female groups at the C5-C6 level. A new parameter, the ratio between translation and rotation, was also established and may prove useful for clinical diagnoses. This parameter has a smaller error associated with it than do pure translations and may aid the clinician by helping to account for the large variation in rotatory ranges of motion within the population. This translation/rotation ratio indicated highly significant differences in the lower segments of the cervical spine between gender groups. Key Words: Cervical spine-Functional flexion/extension x-rays-Kinematics-Normal population.

The cervical spine exhibits many different motion patterns. The atlanto-occipital joint (OCC-C 1) acts as the pivot for the flexion/extension motion of the cranium, while allowing only a few degrees axial rotation (5-7, Panjabi MM, et al, unpublished observations). The atlanto-axial complex (CI-C2), on the other hand, has an extremely wide range of axial r o t a t i o n , t h e a v e r a g e maximum a r o u n d 43" (5,7,10,12), coupled with lateral bending. From the axis down (C2-C7), the cervical vertebrae become more like the lower vertebrae, with rotations and translations in all directions. The prominent motion, though, is flexion/extension. This is due to the alignment of the apophyseal joints and the presence of uncinate processes (10). Knowledge of these

complex motion patterns in the cervical spine is necessary for diagnosis of trauma, degeneration, or other pathologies that may arise there. Abnormal motion in flexion/extension has long been a sign of possible instability in the cervical spine and an indication of the need for surgical interventions, such as fusion. A number of radiographic motion studies using different methods of examination and x-ray analysis have documented this motion in order to facilitate diagnosis of such instabilities. These include Bakke (l), De Seze et al. (3), Buetti-Baumel (2), and Penning (9). They all gathered data from normal subjects using active examinations and measured segmental rotations graphically. Dvorak et al. (4) compared active and passive examination methods using Penning's method of measurement. It was found that a passive examination was more likely to locate a segmental hypermobility. A computer-assisted method was devel-

Received May 8, 1990; accepted April 9, 1991. Address correspondence and reprint requests to Dr. J. Dvorak a t Department of Neurology, Spine Unit, Klinik Wilhelm Schulthess, Neumunsterallee 3, 8008 Zurich, Switzerland.

828

FLEXIONIEXTENSION OF NORMAL CERVICAL SPINE

829

oped for analysis of the flexionlextension motion of the lumbar spine for both normal subjects (6) and patients. It provided both segmental rotations and translations. Error analysis for rotational calculations used by the computer-assisted method was presented by Panjabi et al. (8), along with equations and analysis for computing centers of rotation. These methods were used in this study to quantify the flexion/extension motion of the cervical spine in a normal population. This should present useful kinematic data, which, in future studies with patients, can be used for diagnostic purposes. Final comparison of the original Penning method to the new computer method should prove their usefulness for this region of the spine.

two images of the vertebra B (one on the flexion view, the other on the extension view) are superimposed. The remaining displacement between the two images of vertebra A then represents its motion relative to vertebra B or the intervertebral motion of vertebral level A-B. There are two ways in which this process can be done. The first, the manual method, involves some type of graphical construction, superimposing the images, drawing lines, and measuring the motion by hand. The second method utilizes a computer to enter and superimpose positions of the vertebrae and calculate the desired results. This is the computer-assisted method, which was used earlier for lumbar spine (6), and is also used here.

METHODS

Computer-Assisted Method

Examined Population

Beginning with the extension view, four lines are drawn tangential to the four sides of the vertebral body of each vertebra. Their intersections provide four corner points (Fig. 1). Specific points are chosen for C1 and C2 (Fig. lB). Then each vertebra on the unmarked flexion view is superimposed over its image on the marked extension view and the four

Forty-four healthy asymptomatic adults were examined. The group consisted of twenty-two women and twenty-two men, ages 23-49 years, mean age 31 years. Technique of Passive Roentgenogram Examination

Subjects stood in an upright position with the left side of their bodies closest to the x-ray film. Using a specially designed fixing stand, the mid-thoracic spine and sternum were held to prevent flexion and extension in the thoracic region of the spine. Shoulders were held as low as possible in order to radiographically visualize the cervical-thoracic junction. The distance between the x-ray tube and the film was 150 cm. All passive examinations were performed by the first author, who was sufficiently protected from x-ray exposure. The examiner placed his left hand on the patient’s head and his right on the patient’s chin. The examiner then induced extension until the patient reached the extent of hislher range of motion or reported discomfort. An x-ray was taken in this position and then the process was repeated for flexion. Details are provided elsewhere (4).

W B

Technique of Measurement

Basic Principles

To determine the segmental motion of vertebra A with respect to vertebra B immediately below it, the

FIG. 1. The marking of vertebral bodies. Four lines are drawn tangent to the vertebral body and the intersection of the lines provide the points for the markers (A). The markers show the specific points digitized on the vertebrae C1 and C2 (9).

J Orthop Res, Vol. 9, No. 6 , 1991

830

J . DVORAK ET AL.

corner points are copied by hand from one to the other. A specially designed computer program, running on a personal computer, and a digitizer were used to acquire data and calculate results from the marked x-rays. The four corner points on each vertebra were digitized four times with an average value stored in the computer. This averaging helped lower the human error in digitization. All of the vertebrae on both views were digitized in such a manner. The computer program mathematically superimposed the corner points from the same vertebra on the two different x-rays (flexion and extension) using the positional data from the digitizer, and calculated the following kinematic parameters of spinal motion from extension to flexion (Fig. 2): RX, segmental vertebral rotation in the sagittal plane; AZ, AY, translation of the posterior inferior body corner, or point A, of the upper vertebra along the anterior-posterior and superior-inferior directions, respectively; BZ, BY, translation of the posterior superior body corner, or point B, of the upper vertebra along the anterior-posterior and superiorinferior directions, respectively; and CRZ, CRY, location coordinates of the center of rotation of the upper vertebra from the origin of a coordinate system located at the posterior inferior body corner of the lower vertebra.

The above nomenclature refers to a coordinate system (12) with its origin at the posterior inferior body corner of the lower vertebra. The positive Zaxis is directed anteriorly along a line from the posterior inferior body corner to the anterior inferior body corner of the lower vertebra. The positive Yaxis is directed superiorly and the positive X-axis is directed towards the left. A clockwise rotation, as seen from the origin along a positive axis, is considered positive. Thus, flexion is a positive rotation about the X-axis. In addition, a pair of compound parameters of translation to rotation ratios were computed, similar to those used in the lumbar spine (11). These were AZR = AZIRX and BZR = BZ/RX. The x-ray films in our study were marked using a ruler and a fine ball-point pen. They were then digitized using a Digikon DK2020 digitizer (Kontron, Eching-Munich, West Germany) with a resolution of 0.02 mm. Marking the x-rays took -15 min and digitizing took another 5 min. Statistical Analysis

For each kinematic parameter, differences between vertebral levels were analyzed statistically using analysis of variance (ANOVA) statistics. Differences were significant for p < 0.05. The male/ female gender-related groups were further tested using a Student t-test. Differences were significant for values p < 0.05. Measurement Errors

A single x-ray pair was measured five times, with the lines redrawn and redigitized each time. Standard deviations for five separate digitizations of the same x-ray pair are shown in Table 1 for each of the TABLE 1. Errors, as defined by standard deviation, for five repeated measurements of an x-ray pair of a 32-year-old woman. Rotations in degrees and translations and coordinates of centers of rotation in mm RX

FIG. 2. The coordinate system and the kinematic parameters evaluated in the present study.

J Orthop Res, Vol. 9 , N O . 6 , 1991

Level

SD

A2 SD

AY SD

BZ SD

BY SD

CRZ SD

CRY SD

CUC2 C2iC3 C3/C4 C4iC5 C5/C6 C6/C7

0.93 0.91 0.61 0.63 0.67 0.81

0.34 0.10 0.29 0.21 0.48 0.33

0.66 0.28 0.25 0.20 0.21 0.29

0.24 0.34 0.20 0.32 0.32 0.56

0.57 0.47 0.35 0.20 0.26 0.44

0.92 1.75 0.37 0.29 0.38 0.45

1.26 1.56 0.65 0.79 0.91 0.59

Average

0.76

0.29

0.31

0.33

0.38

0.70

0.97

FLEXIONIEXTENSION OF NORMAL CERVICAL SPINE

motion parameters at all levels. The average values show that the translation measurements, directly extracted from the digitized positions of the vertebra, all deviate -0.3 mm. This value can be assumed to be the result of redrawing the lines around the vertebra for each digitization, as well as human error in the digitization process. Calculated values have the inherent errors of their calculation process and are relatively higher, with RX deviating 0.76 degrees and the center of rotation coordinates differing 0.69 mm in the Z-direction and 0.96 mm in the Y-direction. Differences in values between levels can be attributed to x-ray quality, degree of repeatability, and specific calculation considerations for that specific level. As expected, Cl-C2, C2-C3, and C6-C7 have generally larger variations than the other middle levels. RESULTS

A healthy population of 44 asymptomatic adults (22 women and 22 men) was examined. Means and standard deviations for each kinematic parameter are presented in Table 2. Also indicated are the statistically significant differences between levels in the gender groups. Rotations

Mean values of the segmental rotations are shown in Fig. 3 by means of functional diagrams. Results show the mean values plus and minus two standard deviations for male and female groups. It was found that in the level C5-C6, a significant difference (p < 0.05) in the gender-related groups was present. In the adjacent levels of C4-C5 and C6-C7, a trend toward significance was present as well (p TABLE 2. Means and standard deviations of the motions Segments AZ(mm) AY (mm) BZ(mm) BY (mm)

CRZ(mm) CRY (mm)

C 1 4 2 C2-C3 -3.8 1.6 6.2 2.3 -1.4 1.4 8.1 3.0 -4.1 4.2 30.0 5.6

2.4 0.9 1.8 0.8 6.9 1.7 3.0 1.3 4.0 3.5 9.4 4.8

C3-C4

CkC5

3.2 1 .o 2.3 1 .o 8.5 1.8 3.6 1.2 4.3 2.7 9.7 3.4

3.6 1.2 2.9 0.7 10.0 1.9 4.2 1.o 6.0 2.2 10.4 2.8

C5-C6

C6C7

2.9 1.1 3.2 0.8 9.8 1.9 4.3 1.0 6.4 1.8 12.9 2.5

2.0 0.9 3.1 0.8 8.4 1.9 3.9 0.9 6.4 2.5 17.2 2.1

831

= 0.059 and p = 0.054, respectively). In general, for the entire population, it can be seen that rotations of the lower three vertebrae were similar, with C5-C6 having the largest value. The amount of rotation decreased from C5-C6 until it reached a minimum at C2-C3. The rotation of Cl-C2 was slightly greater than this value but it should be noted that this value had a much larger spread of values than did other segments, as seen by its large standard deviation.

Translations

Mean values for the translations of point A and point B are also shown in Table 2. A functional diagram of BZ with means and plushinus two standard deviations is shown in Fig. 4. These values were more difficult to compare to each other due to variations in the anatomy of the cervical spine. The lower vertebrae followed the same pattern of translation as they did in rotation, increasing in value moving down the spine with C4-C5 exhibiting the most translation in the Z-direction and C5-C6 the most in the Y-direction. Segment C6-C7 showed an average amount of translation in the Y-direction but markedly smaller values of translation in the Zdirection compared to the other lower segments. In comparing the gender-related groups, it was found that a significant difference was present for both translations in the Y-direction, AY and BY; however, translations in the Z-direction were not significantly different across the gender groups. Centers of Rotation

The two-dimensional coordinates of the segmental centers of rotation are also shown in Table 2, but are more readily understood by viewing Fig. 5. From C6-C7 to C2-C3, the centers of rotation move lower for each successive segment. For C6-C7, the center of rotation is located on the superior endplate, and for C2-C3, it is located in the middle of body. It also gradually moves posteriorly. The center of rotation of Cl-C2 is located near the position of the transverse portion of the cruciate ligament. Translation/Rotation Ratio

Mean values and standard deviations for AZR and BZR are also given in Table 3. For the genderrelated groups, the ratio of BZ to RX showed a high

J Orthop Res, Vol. 9, No. 6 , 1991

832 A

35

J . DVORAK ET AL. j

-

Average

0C1-C2

C2-C3

C3C4

C4-CS

C6-C7

C5-C6

Spine Segment

B

35

7

-

10

* - ...........................................................................................................................................................................

.......................

A

0-

FIG. 3. Functional diagram of the sagittal plane rotation, RX, showing the mean values plus and minus two standard deviations for both gender groups, (A) male flexion/extension, (B) female flexion/ extension.

Avg +2 SD Average Avg-2SD

............................

. ..

I

c1-c2

C2-C3

c3-c4

c4-c5

C5-C6

C6-C7

Spine Segment

10 --

L

E

E

FIG. 4. Functional diagram of the flexion/ extension sagittal plane translation (B2) of the anterior-posterior translation of the upper posterior corner of the vertebral body, TBZ, showing the mean values and plus/minus two standard deviations.

0

--t-

.

-5

- '

Avg+ZSD Average Aq-2SD

_ I I

-10 c 1- c 2

C2-C3

c3-c4

c4-c5

Spine Segment

J Orthop Res, Vol. 9, No. 6 , 1991

C5-C6

C6X7

FLEXIONIEXTENSION O F NORMAL CERVICAL SPINE

833

TABLE 3 . Translationhotation ratio. The mean values for the translationlrotation ratios of points A and B for each vertebral level of the cervical spine, their standard deviations and the SDlmean ratio Level

C1/C2 level

C2/C3 level

C3/C4 level

C4/C5 level

CWC6 level

CfX7 level

AZIRX Cl-C2 C2-C3 c3-c4 CK5 C5-C6 ccc7 Average Male values Level BZ/RX C1-C2 C2-C3 c3-c4 CK5 C5-C6 ccc7 Average Female values Level BZ/RX c 1-c2 C2-C3 c3-c4 C K 5 (25x6 ccc7 Average

Mean

SD

-0.28 0.20 0.19 0.17 0.13 0.09

0.09 0.08 0.06 0.05 0.04 0.03

34% 40% 31% 30% 33% 33% 34%

Mean

SD

SD/rnean

- 0.10

0.11 0.09 0.06 0.07 0.04 0.05

107% 15% 11% 14% 9% 12% 28%

Mean

SD

SD/rnean

-0.11 0.56 0.47" 0.45" 0.41" 0.36"

0.11 0.07 0.05 0.06 0.04 0.04

100% 13% 11% 13% 10% 11% 26%

0.59 0.53" 0.50" 0.46" 0.41'

S D h e an

Values are in rnrn per degree rotation.

FIG. 5. The positions of the centers of rotations for each motion segment. The marker shows a mean position and the gray dashed line encloses the area of two standard deviations of this mean. Positions are marked as center of rotation of level/with respect to level (i.e., C3/C4, COR of C3 with respect to C4).

degree (p < 0.005) of significance in the levels from C3-C4 to C6-C7. DISCUSSION

Since diagnosis of segmental instability of the cervical spine causes common difficulties, a careful investigation of functional x-rays might be useful. There were at least three reasons for determining a new parameter-the translationtrotation ratio-for the two posterior corners A and B of the vertebra. First, the translation/rotation ratio has been advocated a s a potential instability parameter (1 1). Secondly, translation/rotation ratio parameters (AZR and BZR) have less standard deviation ratios than d o corresponding translation parameters, AZ or

" Values for the BZIRX ratio have been given for both gender groups to illustrate the statistically significant difference in levels.

BZ. Thus, adjusting for units of measurement via the standard deviation ratio, we see that AZR and BZR have fewer errors than d o AZ and BZ, respectively. Third, if the translatory measurements are to be used clinically for determination of spinal instability, then some account must be taken of the large variation in the rotatory ranges of motion within the population. Dividing the translation by the rotatory range of motion at the same level compensates, to a certain extent, for this variation. Our results have proven the validity of the last two statements. For example, the average (over all the cervical levels) standard deviation ratio for AZR was 34% compared to 39% for the AZ, and the corresponding values for the BZR and BZ were 27% (28% men, 26% women) and 34%, respectively. However, the real value of this parameter is not definitively proven and should be confirmed in patient population studies. The BZR with its lower standard de-

J Orthop Res, Vol. 9, No. 6 , 1991

J . DVORAK ET AL. viation ratio may be the parameter of choice in determining spinal instability. The fact that the BZR parameter was significantly different for the lower segments of the cervical spine between male and female populations is interesting to note. While the importance of this difference in not yet understood, it indicates that more study in this area is needed and, in fact, an age-group/gender related study is currently underway to further address this correlation. Results of centers of rotation of the middle and lower cervical vertebrae were similar to those of Penning (9), showing the same changes in the position of the center of rotation along the cervical spine. These changes in position can be directly related to the type of motion exhibited by the vertebra from extension to flexion. For the lower vertebrae, the motion is more of a “tilting” motion, as in the lumbar spine, while the upper vertebrae exhibit a “gliding” motion. A gliding motion requires greater translation than does a tilting one for the equivalent amount of rotation. Thus, gliding motion is more unstable than tilting motion (10). These gliding segments, such as C2-C3, require smaller translations and rotations, as compared to the tilting segments, in order to maintain stability. Penning hypothesized that the larger uncinate processes of these upper segments added the additional stability necessary for this type of motion. Tilting motion is most clearly exhibited by C6-C7, where the segment’s rotation is as large as that of neighboring segments, but the translations are smaller. A tilting motion could be evidence that a segment carries an increased load, as in the lumbar spine. The mean center of rotation of Cl-C2 is located near the position of the transverse portion of the cruciate ligament. This ligament acts to prevent abnormal anterior translation, such as would occur in moving from an extended to a flexed position, as in our experiment. So, by restraining anterior translation, the transverse ligament also acts as the pivot about which C1 rotates. Again, the anatomical differences and x-ray resolution combine to give a wide variation of values.

J Orthop Res, Val. 9, No. 6 , 1991

Future studies should include a survey of patient populations with diagnosed pathologies, looking at the differences in all parameters as related to gender and age. It can then be assessed if the technique of functional flexion/extension x-rays will be useful in differentiating between and locating specific pathologies in patients with cervical spine disorders, such as soft tissue injury (i.e., whiplash injury), radicular syndromes, and degenerative changes. Acknowledgment: This study was supported by the research fund of the W. Schulthess Hospital.

REFERENCES 1 . Bakke S: Roentgenologische Beobachtungen Ueber die Bewegungen der Halswirbelsaule. Acta Radio1 13;suppl:1-68, 1931 2. Buetti-Bauml C: Funktionelle Roentgendiagnostik der Halswirbelsaule. Thieme, Stuttgart, Fortschritte auf dem Gebiete der Roentgenstrahlen vereinigt mit Roentgenpraxis. Erganzungsband 70: 1-1 12, 1954 3. De Seze C, Dijian A, Abdelmoula M: Etude Radiologique de la Dynamique cervical dans la plaine sagittal. (Une contribution radiophysiologique a l’etude pathogenique des artheoses cervicales). Rev Rhum Ma1 Osteoartic 18:3746, 1951 4. Dvorak J, Froelich D, Penning L, Baumgaertner H, Panjabi MM: Functional radiographic diagnosis of the cervical spine: flexioniextension. Spine 13:748-755, 1988 5 . Dvorak J, Hayek J, Zehnder R: CT-functional diagnostics of the rotatory instability of the upper cervical spine part 2. an evaluation on healthy adults and patients with suspected instability. Spine 12:726-731, 1987 6. Dvorak J, Panjabi MM, Chang DG, Theiler R, Grob D: Functional radiographic diagnosis of the lumbar spine: flexiodextension and lateral bending. Spine (in press) 7. Dvorak J, Panjabi MM, Gerber M, Wichmann W: CTfunctional diagnostics of the rotatory instability of upper cervical spine: 1 . an experimental study on cadavers. Spine 12:197-205, 1987 8. Panjabi MM, Goel VK, Walter SD, Schick S: Errors in the center and angle of rotation of a joint: An experimental study. J Biomech Eng 104:232-237, 1982 9. Penning L: Functioneel rontgenonderzoek Bij degenerative en traumatische afwijkingen der laag-cervicale Bewegingssegmenten [Thesis]. University of Gronikge, The Netherlands, 1960 10. Penning L: Review paper: Differences in anatomy, motion, development and aging of the upper and lower cervical disk segments. Clin Biornech 3:3747, 1988 1 1 . Stokes IAF, Frymoyer JW: Segmental motion and instability. Spine 12:688-691, 1987 12. White AA, Panjabi MM: Clinical Biomechanics ofthe Spine. Philadelphia: JB Lippincott Company, 1978