The Spine Journal 9 (2009) 711–714

Clinical Study

The prevalence cervical facet arthrosis: an osseous study

in a cadveric population
Michael J. Lee, MDa,*, K. Daniel Riew, MDb
a
Department of Orthopaedic Surgery, University of Washington Medical Center, 1959 Pacific St NE, Box 356500, Seattle, WA 98195, USA
b
Department of Orthopaedic Surgery, Barnes-Jewish Hospital at Washington University School of Medicine, St Louis, MI, USA
Received 24 May 2008; accepted 17 April 2009

Abstract BACKGROUND CONTEXT: Cervical facet arthrosis has been implicated as a cause for neck
pain, radiculopathy, occipital headache, and ear pain.
PURPOSE: The objective of this study was to examine the occurrence of facet arthrosis in the
cervical spine.
STUDY DESIGN/SETTING: This study examined cadaveric specimens from the Hamann Todd
Collection.
PATIENT SAMPLE: None.
OUTCOMES MEASURES: None.
MATERIALS AND METHODS: Four hundred sixty-five skeletally mature human cervical spines
from the Hamann Todd Collection in the Cleveland Museum of Natural History were obtained for anal-
ysis. We analyzed the facets for arthrosis. We graded no arthrosis as Grade 0. Facets with peripheral os-
teophytic reaction, but with no lateral mass distortion were graded as Grade 1. Facets with peripheral
osteophytic reaction and lateral mass distortion were graded as Grade 2. Facets that were ankylosed were
graded as Grade 3. Each specimen was examined bilaterally at levels from C2–C3 through C6–C7, yield-
ing 4,650 specimen assessments. The data were analyzed to compare cervical levels, gender, facet side,
age groups, and race. Proportion analysis, using the Fisher exact test, was used to assess for statistical
difference between various groupings.
RESULTS: In the entire population of 465 specimens, the upper cervical specimens appeared to be
affected by facet arthrosis more frequently than the lower levels; 12.37% of the specimens had bony
evidence of arthrosis at the C2–C3 level; 13.33% of the specimens had arthrosis occur at the C3–C4 level;
14.62% at the C4–C5 level; 7.85% at the C5–C6 level, and 4.84% at the C6–C7 level. The large majority
of all cervical facet arthrosis was found to be Grade 1 at all levels. In the older population, the prevalence
of facet arthrosis is as high as 29.87% for the C4–C5 level. C4–C5 level appears to be affected the most
frequently, followed by the C3–C4 level, then C2–C3, C5–C6, and C6–C7.
CONCLUSION: The prevalence of cervical facet arthrosis increases with age, and occurs more
commonly in the upper cervical spine. Ó 2009 Elsevier Inc. All rights reserved.
Keywords: Cervical facet; Arthrosis; Cervical spine

Introduction there is much literature describing the spondylosis pathol-

ogy anteriorly, there are few descriptions of cervical facet
Spondylosis is a frequent cause of neural tissue com-
arthrosis in the literature. The purpose of this study was
pression in the cervical spine. Cervical facet arthrosis can
to determine the prevalence of osseous cervical facet
occur and osteophytic reaction at the superior facet can
arthrosis in a cadaveric population.
cause encroachment on the foramen posteriorly. Although
356500, Seattle, WA 98195, USA. Tel.: (206) 543-3690; fax: (206) 685-
FDA device/drug status: not applicable.
3139.
Author disclosures: KDR (staff and materials, Medtronic).
E-mail address: jihoon2000@hotmail.com (M.J. Lee)
* Corresponding author. Department of Orthopaedics and Sports Med-
icine, University of Washington Medical Center, 1959 Pacific St NE, Box

1529-9430/09/$ – see front matter Ó 2009 Elsevier Inc. All rights reserved.
doi:10.1016/j.spinee.2009.04.016
712 M.J. Lee and K.D. Riew / The Spine Journal 9 (2009) 711–714

Table 2
Grading scale of facet arthrosis
Grade
0 No evidence of bony arthrosis
Context 1 Peripheral osteophytic lipping; Fig. 1
Cervical facet arthrosis may result in significant clinical no lateral mass changes
symptoms. This article assesses its prevalence using ca- 2 Peripheral osteophytic changes Figs. 2 and 3
with lateral mass distortion
deveric skeletal specimens.
3 Ankylosis
Contribution
The authors have found that cervical facet arthrosis in-
creases with age and occurs more commonly in the up- Results
per cervical levels.
In the entire population of 465 specimens, the upper
Implications cervical specimens appeared to be affected by facet arthro-
By letting us know just how common a particular path- sis more frequently than the lower levels. The results are
ologic process is, there is an academic value to preva- summarized in Table 3. The large majority of all cervical
lence studies. The practical value, however, rests upon facet arthrosis was found to be Grade 1 at all levels (Figs.
the ability of the information to provide pre-test proba- 4 and 5). In the oldest group (O70 years), the prevalence of
bilities for specificity of diagnostic testing. Thus, while facet arthrosis was as high as 29.9% for the C4–C5 level
the classic cadaveric studiesdfrom Wood-Jones, at the (Table 4). The left side had a higher rate of facet arthrosis
turn of the 20th century, to the current ones using the at all levels, but was not statistically significant (Table 5).
Todd collectiondadd elegance to the literature, tying No significant difference was observed between gender
such prevalence to symptom causation and mechanism groups or racial groups and the occurrence of facet
of disease is vital. Yet this has been proved difficult in arthrosis.
many cases, especially as it pertains to axial neck/back
pain.
dThe Editors
Discussion
Materials and methods Cervical spondylosis including anterior and posterior de-
generative changes has been extensively discussed in the
We obtained 465 cervical spine skeletal specimens from
current literature. However, there has been little attention
the Hamann Todd Collection at the Cleveland Museum of
specifically examining the prevalence of cervical facet ar-
Natural History. The Hamann Todd Collection is an osteo-
throsis. Cervical facet arthrosis been identified as a cause
logical collection of fully disarticulated human skeletons
for radiculopathy, neck pain, and more rarely, myelopathy,
and has been used extensively for research purposes. Age,
ear pain, and occipital headache [1–5]. This study specifi-
gender, and race distribution are summarized in Table 1.
cally examines the prevalence and severity of cervical facet
Each specimen was examined for the presence of bony
arthrosis in cadaveric population.
changes reflective of facet arthrosis. There is no grading
scale for facet arthrosis, so one was devised for the purpose
of this study (Table 2; Figs. 1–3).
Each specimen was examined bilaterally at levels from
C2–C3 through C6–C7, yielding 4,650 specimen assess-
ments. The data were analyzed to compare cervical levels,
gender, facet side, age groups, and race. Proportion analysis
(chi-square and Fisher exact) was used to assess for statis-
tical difference between various groupings.

Table 1
Distribution of age, gender, and race in study population
Age at death (y) Gender Race
30–39 77 Male 391 Caucasian 328
40–49 92 Female 74 African American 137
50–59 115
60–69 111
Fig. 1. Lateral view of Grade 1 arthrosis. Peripheral lipping without lat-
O70 77
eral mass distortion.
 M.J. Lee and K.D. Riew / The Spine Journal 9 (2009) 711–714
Fig. 2. Lateral view of Grade 2 arthrosis. Osteophytic reaction with lateral
mass distortion.
The present study is a large cadaveric study of osseous
facet arthrosis. Cervical facet arthrosis is rare in people
younger than 40 years of age. The prevalence expectedly
rises over time. We observed that almost 30% of people
older than 70 years had bony evidence of facet arthrosis.
Spondylosis most commonly affects the cervical spine at
the C5–C6 level. In this study, however, facet arthrosis most
commonly affected the upper cervical levels. C4–C5 was
the most commonly affected level, followed by C3–C4
and C2–C3. This finding is supported by previous observa-
tion. Friedenberg et al. dissected 41 human cervical speci-
mens and noted that the prevalence of facet arthrosis in the
cervical spine was highest at C3–C4 (17%, seven speci-
mens) [6]. The severity of facet arthrosis observed in this
study was most severe at C2–C3. In a separate study, Frie-
denberg and Miller observed that highest rate of facet ar-
throsis appeared to be at the C4–C5 level by assessment
through clinical X-rays [7]. Lestini and Wiesel, in a review,
Fig. 3. Axial view of Grade 2 arthrosis. Osteophytic reaction with lateral
mass distortion.
713
Fig. 4. Arthrosis: grade versus level.
noted that facet arthrosis was observed more commonly in
the upper cervical spine than in the lower cervical spine [8].
Though this finding has been previously observed, the eti-
ology of this finding is unclear. Freidenberg et al. postu-
lated that this was likely because of the shift in weight in
the upper cervical spine. In a normal lordotic cervical
spine, the weightbearing axis is shifted more posteriorly
toward the facets in the upper cervical spine, which may
account for this finding.
This study does have limitations. Firstly, this was a cadav-
eric study examining 465 specimens. There is no informa-
tion correlating to the presence or absence of symptoms.
Although osseous facet arthrosis can be observed and
graded, it is impossible to draw definitive clinical conclu-
sions based on this study alone. Secondly, because this is
Fig. 5. Arthrosis for age and level.
714 M.J. Lee and K.D. Riew / The Spine Journal 9 (2009) 711–714

Table 3 Table 5
Incidence of facet arthrosis at each level Facet arthrosis in right versus left sides
All patients, n5465 C2–C3 C3–C4 C4–C5 C5–C6 C6–C7 % With arthrosis
% Grade 1 8.82 10.11 11.72 5.48 3.33 Side C2–C3 C3–C4 C4–C5 C5–C6 C6–C7
% Grade 2 0.97 2.15 2.04 0.43 0.22 Right facet 10.97 11.18 12.90 6.45 4.73
% Grade 3 2.58 1.08 0.86 1.94 1.29 Left facet 13.76 15.48 16.34 9.25 4.95
% With 12.37 13.33 14.62 7.85 4.84
any arthrosis

observed at C5–C6 and C6–C7, cervical facet arthrosis ap-

Table 4 pears to be more common in the upper cervical spine at
Arthrosis for age and level
C3–C4 and C4–C5, and its prevalence increases with age.
% Arthrosis
Age C2–C3 C3–C4 C4–C5 C5–C6 C6–C7
30–39 y 0.00 0.00 2.14 1.43 1.43 References
40–49 y 11.41 10.33 7.61 5.43 4.89
50–59 y 11.74 12.61 10.00 7.83 5.65 [1] Houser OW, Onofrio BM, Miller GM, Folger WN, Smith PL,
60–69 y 16.52 17.55 14.87 6.31 4.05 Kallman DA. Cervical neural foraminal canal stenosis: computerized
O70 y 25.24 27.95 29.87 19.49 9.09 tomographic myelography diagnosis. J Neurosurg 1993;79:84–8.
[2] Lamer TJ. Ear pain due to cervical spine arthritis: treatment with
cervical facet injection. Headache 1991;31:682–3.
an osseous cadaveric study, the occurrence of facet arthrosis [3] Connell MD, Wiesel SW. Natural history and pathogenesis of cervical
disk disease. Orthop Clin North Am 1992;23:369–80. Review.
in this study is likelihood underestimated. We are unable
[4] Santavirta S, Konttinen YT, Lindqvist C, Sandelin J. Occipital head-
to assess for cartilage loss within the facet, capsular attenu- ache in rheumatoid cervical facet joint arthritis. Lancet 1986;2:695.
ation, or other soft issue pathology, which may contribute to [5] Omura K, Hukuda S, Matsumoto K, Katsuura A, Nishioka J,
symptoms. Imai S. Cervical myelopathy caused by calcium pyrophosphate di-
hydrate crystal deposition in facet joints. A case report. Spine
1996;21:2372–5.
[6] Friedenberg ZB, Edeiken J, Spencer HN, Tolentino SC. Degenerative
Conclusion changes in the cervical spine. J Bone Joint Surg Am 1959;41-A:61–70.
[7] Friedenberg ZB, Miller WT. Degenerative disc disease of the cervical
The prevalence of cervical facet arthrosis increases with spine. J Bone Joint Surg Am 1963;45:1171–8.
age, and occurs primarily in the upper cervical spine. [8] Lestini WF, Wiesel SW. The pathogenesis of cervical spondylosis.
Although cervical spine spondylosis is most commonly Clin Orthop Relat Res 1989;239:69–93.