engleski

Bone distribution in the cross sections throughout the femoral neck and intertrochanteric regions: a study using clinical QCT

Recent data (Mayhew et al, 2005, Lancet 366: 129) supported the suggestion that the buckling of the posterior part of the superolateral cortex of the femoral neck (FN) results in hip fracture during sideway fall. The purpose of this study is to see if clinical QCT can be used to study this topic. We analysed the baseline-visit CT scans of both proximal femurs of 27 women (mean age 81, range 65-86 yr) with osteoporosis who were recruited at a single centre participating in the HORIZON Study. The CT number was converted to density (r) using a phantom. The maximum r (rmax) of each femur was determined and the voxel bone volume fraction (BVF) defined as the voxel r/rmax. Cortical bone was assumed with BVF>=0.5. A cortical thickness of 0.5 mm was imposed, by assigning BVF to 0.5, on surface voxels with original BVF<0.5. The femurs were digitally rotated to a standard orientation to ensure a consistent cross section (CS) definition. The FN and inter-trochanteric (IT) axes were determined and the CSs defined along the axes with a spacing of about 1.5 mm. Each CS was divided into 8 sectors with the centre at the bone mass centre: Inferior, Infero-Anterior, Anterior, Supero-Anterior, Superior, Supero-Posterior, Posterior, Infero-Posterior (Figure). Bone area (BA), cortical shell thickness (CST) and moments of inertia (MI) were calculated for the whole CS and each octant. The results presented here relate to CSs at the start, middle and end of FN, middle and end of IT, and minimum MI about the antero-posterior axis in the FN and IT. One-way ANOVA of the middle FN data showed that the largest BA and CST were in octant I, and minimum BA and CST were in octant S-P and P respectively. Two-way ANOVA of the data showed that CS and octant position and their interaction were significant factors affecting the calculated parameters. However, if the parameters were expressed as the % of their CS values, the CS position was no longer significant. Scheffe post-hoc comparison revealed that the octant I had the largest BA and CST, while octant A and P had the lowest BA and CST respectively. In conclusion, this study used clinical QCT and was able to confirm the findings of micro CT on excised samples: the thinnest part of the proximal femur is the posterior octant.

Bone distribution; femoral neck; intertrochanteric region; QCT

nije evidentirano