INTRODUCTION: Low-energy pelvic fractures have been increasing in the elderly with a 2-fold rise in the occurrence of acetabular fractures in the last three decades [1]. Acetabular support, orientation and coverage parameters have been identified to correlate with over-coverage, hip dysplasia and hip osteoarthritis [2]. However, structural and biomechanical factors that contribute to low-energy acetabular fractures are unknown. Therefore, we aimed to clarify the three-dimensional (3D) morphometric upper femur and acetabular parameters that contribute to acetabular fractures. 

METHODS: The data consisted of abdominopelvic computed tomography (CT) images of 120 acetabular fracture patients (31 female, 89 male, mean age 72 years, standard deviation 12 years) and 120 age and gender matched healthy controls. Ethical permission was obtained from the Ethical Committee of Northern Ostrobothnia Hospital District, Oulu University Hospital. The pelvic bones and femurs in CT images were thresholded, segmented and modeled in 3D using Materialise Mimics software. The model was reoriented into the anatomical coordinate system and the morphological parameters were measured semi-automatically using Materialise 3-Matic software. The upper femur parameters included Neck Shaft Angle (NSA), Femoral Neck Length (FNL) and Neck Shaft Width (NSW), and the acetabular support/coverage parameters included Acetabular Inclination (AI), Acetabular Anteversion (AAnt), Acetabular 
Angle of Sharp (AA), Lateral Center Edge Angle (LCEA), Anterior Center Edge Angle (ACEA), Posterior Center Edge Angle (PCEA), Anterior Sector 
Angle (ASA), and Posterior Sector Angle (PSA). As the parameters were normally distributed, an independent samples t-test was performed to compare the differences between fracture cases (contra-lateral side) and controls. Pearson correlation analysis was performed to assess the relation of the measured parameters within fracture cases and control cases. 

RESULTS: NSA was statistically significantly lower (p<0.05) in fracture cases than in controls. This difference was also significant for both females and males. There was no statistically significant difference in FNL and NSW between fracture and control groups and within genders. AA was significantly lower in fracture cases compared to controls. However, AA was only significant in females but not in males. The rest of the acetabular support/coverage parameters (LCEA, AI, AAnt, ACEA, PCEA, PSA and ASA) did not differ between fracture and control groups (Table 1).  Furthermore, even though the mean differences were not significant there were some gender difference in AA and PSA. There was a significant correlation (p<0.05) between NSA and LCEA (R=-0.24), NSA and ASA (R=-0.24), NSW and AA (R=-0.30), and between NSW and LCEA (R=-0.25) amongst the fracture cases as opposed to the controls that showed no significant correlation between these parameters. In the control group there was a significant correlation (p<0.01) between FNL and AAnt (R=-0.441), FNL and PCEA (R=-0.52), FNL and PSA (R=-0.43), NSW and AAnt (R=-0.36), NSW and PCEA(R=-0.43), NSW and PSA (R=-0.40).  

DISCUSSION: The current findings suggest that smaller NSA is indicative of acetabular fractures. Previously, larger NSA has been shown to be associated with femoral neck fractures [3], [4]. The values of AA and LCEA indicate over-coverage for both control and fracture groups [5]. However, those with acetabular fractures were severely over-covered as indicated by the smaller AA. The relationship of NSA to either acetabular fractures or femoral head fractures may be due to the physical orientation that the femur has during low-energy impacts which results in unique stress/strain distributions. Finally, the association between NSA, femoral head over-coverage and acetabular fractures might be explained by the slightly low significant correlation between NSA and LCEA mainly because it was found only in the fracture groups. It is worth noting here that no significant association was observed between NSA and AA in both groups, suggesting that these parameters are measuring independent phenomena and individually contribute to the acetabular fracture. In future, it should be clarified that if in fact NSA and AA are independent factors or whether their interaction contribute to acetabular fractures.  

SIGNIFICANCE/CLINICAL RELEVANCE: The findings of this study form basis for understanding the structural and biomechanical conditions that 
lead to the occurrence of low-energy acetabular fractures in the elderly. Clinically, identifying the abnormalities in the morphology of the femur and/or pelvis can lead to better diagnosis and treatment options. 

REFERENCES: [1] N. M. Beckmann et al., Emerg. Radiol., 531–539, 2017. [2] H. van Bosse et al., Clin. Orthop. Relat. Res., 1712–1723, 2015. [3] J. 
Thevenot et al., Radiology, 184–191, 2014. [4] P. Pulkkinen et al., Osteoporos. Int., 1269–1276, 2010. [5] M. Tannast et al., Clin. Orthop. Relat. Res.,  1234–1246, 2015.

ACKNOWLEDGEMENTS: The study was supported by the CINOP NICHE project ETH246

1581 - Three-Dimensional Computed Tomography Based Morphometric Study of Low-Energy Acetabular Fractures