Biomechanical and anatomical considerations in lumbar spinous process fixation--an in vitro human cadaveric model.

Background Context: Although multiple mechanisms of device attachment to the spinous processes exist, there is a paucity of data regarding lumbar spinous process morphology and peak failure loads.

Purpose: Using an in vitro human cadaveric spine model, the primary objective of the present study was to compare the peak load and mechanisms of lumbar spinous process failure with variation in spinous process hole location and pullout direction. A secondary objective was to provide an in-depth characterization of spinous process morphology.

Study Design: Biomechanical and anatomical considerations in lumbar spinous process fixation using an in vitro human cadaveric model.

Methods: A total of 12 intact lumbar spines were used in the current investigation. The vertebral segments (L1-L5) were randomly assigned to one of five treatment groups with variation in spinous process hole placement and pullout direction: (1) central hole placement with superior pullout (n=10), (2) central hole placement with inferior pullout (n=10), (3) inferior hole placement with inferior pullout (n=10), (4) superior hole placement with superior pullout (n=10), and (5) intact spinous process with superior pullout (n=14). A 4-mm diameter pin was placed through the hole followed by pullout testing using a material testing system. As well, the bone mineral density (BMD) (g/cm(3)) was measured for each segment. Data were quantified in terms of anatomical dimensions (mm), peak failure loads (newtons [N]), and fracture mechanisms, with linear regression analysis to identify relationships between anatomical and biomechanical data.

Results: Based on anatomical comparisons, there were significant differences between the anteroposterior and cephalocaudal dimensions of the L5 spinous process versus L1-L4 (p<.05). Statistical analysis of peak load at failure of the four reconstruction treatments and intact condition demonstrated no significant differences between treatments (range, 350-500 N) (p>.05). However, a significant linear correlation was observed between peak failure load and anteroposterior and cephalocaudal dimensions (p<.05). Correlation between BMD and peak spinous processes failure load was approaching statistical significance (p=.08). 30 of 54 specimens failed via direct pullout (plow through), whereas 8 of 54 specimens demonstrated spinous process fracture. The remaining cases failed via plow through followed by fracture of the spinous process (16 of 54; 29%).

Conclusions: The present study demonstrated that variation in spinous process hole placement did not significantly influence failure load. However, there was a strong linear correlation between peak failure load and the anteroposterior and cephalocaudal anatomical dimensions. From a clinical standpoint, the findings of the present study indicate that attachment through the spinous process provides a viable alternative to attachment around the spinous processes. In addition, the anatomical dimensions of the lumbar spinous processes have a greater influence on biomechanical fixation than either hole location or BMD.