Objective: To evaluate squaring of cervical spine in ankylosing spondylitis (AS) and to suggest whether assessment for squaring of entire cervical spines in the modified Stoke AS Spinal Score (mSASSS) is really needed.
Methods: The study group comprised 125 healthy subjects and 135 patients with AS. Two experienced radiologists assessed the presence or absence of squaring in the individual cervical vertebrae twice independently. The frequencies of squaring of each vertebral site were obtained, and the differences between the control and AS groups were statistically analyzed.
Results: Kappa intra- and interobserver reliability coefficients were 0.76 and 0.61, respectively. The frequency of spinal squaring for each vertebra in the control group was as follows: the 2nd (lower border 100%), 3rd (upper border 85.2%-91.0%, lower border 91.0%-93.4%), 4th (upper border 44.3%-49.2%, lower border 57.4%-63.1%), 5th (upper border 5.0%-6.6%, lower border 8.4%-15%), 6th (upper border 2.7%-7.3%, lower border 4.7%-8.6%), 7th cervical vertebra (upper border 39.7%-45.6%, lower border 51.5%-54.1%), and the 1st thoracic vertebra (upper border 56.0%-62.5%). Although there were statistically significant differences at 4th, 5th, 6th, and upper border of 7th cervical vertebra, a high proportion of normal 4th and 7th cervical vertebra were scored as squared.
Conclusion: Only the 5th and 6th cervical spines are useful for scoring squaring of cervical spine in the mSASSS.