Bs. Myers et al., THE ROLE OF IMAGING AND IN-SITU BIOMECHANICAL TESTING IN ASSESSING PEDICLE SCREW PULL-OUT STRENGTH, Spine (Philadelphia, Pa. 1976), 21(17), 1996, pp. 1962-1968
Study Design. This study determined the predictive ability of quantita
tive computed tomography, dual energy x-ray absorptiometry, pedicular
geometry, and mechanical testing in assessing the strength of pedicle
screw fixation in an in vitro mechanical test of intrapedicular screw
fixation in the human cadaveric lumbar spine. Objective. To test sever
al hypotheses regarding the relative predictive value of densitometry,
pedicular geometry, and mechanical testing in describing pedicle scre
w pull-out. Summary of Background Data. Previous investigations have s
uggested that mechanical testing, geometry, and densitometry, determin
ed by quantitative computed tomography or dual energy x-ray absorptiom
etry, predict the strength of the screw-bone system. However, no study
has compared the relative predictive value of these techniques. Metho
ds. Forty-nine pedicle screw cyclic-combined flexion-extension moment-
axial pull-out tests were performed on human cadaveric lumbar vertebra
e. The predictive ability of quantitative computed tomography, dual en
ergy x-ray absorptiometry, insertional torque, in situ stiffness, and
pedicular geometry was assessed using multiple regression. Results. Se
veral variables correlated to force at failure. However, multiple regr
ession analysis showed that bone mineral density of the pedicle determ
ined by quantitative computed tomography, insertional torque, and in s
itu stiffness when used in combination resulted in the strongest predi
ction of pull-out force. No other measures provided additional predict
ive ability in the presence of these measures. Conclusions. Pedicle de
nsity determined by quantitative computed tomography when used with in
sertional torque and in situ stiffness provides the strongest predicti
ve ability of screw pull-out. Geometric measures of the pedicle and de
nsity determined by dual energy x-ray absorptiometry do not provide ad
ditional predictive ability in the presence of these measures.