Poster Presentation 29th Australian and New Zealand Bone and Mineral Society Annual Scientific Meeting 2019

Differences in femoral neck anatomy in Beijing women with and without femoral neck fractures (#97)

Richard L Prince 1 , Benjamin CC Khoo 2 , Ling Wang 3 , Joshua R Lewis 1 4 , Keenan J Brown 5 , Xiaoguang Cheng 3
  1. University of Western Australia, Crawley, WA, Australia
  2. Medical Technology and Physics, Sir Charles Gairdner Hospital, Nedlands, Western Australia, Australia
  3. Department of Radiology , Beijing Jishuitan Hospital, Beijing, China
  4. School of Medical and Health Sciences, Edith Cowan University, Joondalup, Western Australia, Australia
  5. Mindways Software, Austin, Texas, USA

Fractures of the proximal femur are the most severe osteoporotic fractures. We have previously validated a method of calculating a two dimensional projection from three dimensional Quantitative Computed Tomography (QCT) images using CTXA software (Mindways Software Inc., USA).  Variables include areal bone density (FNaBMD) and femoral neck width (FNW) and two internal measures of bone distribution the standard deviation of mineral-mass projection profile distribution related to bone distribution (FNsigma) and the displacement between center-of-mineral-mass.  We have previously reported improved hip fracture discrimination in the PLSAW study (1).

We now hypothesize that the inclusion of these variables related to  the distribution of bone within the 2D cross section would improve hip fracture discrimination in a case control study of 285 Chinese female FN fracture and 261 community controls mean age 71-72. Structural variables were calculated from QCT scans and analyzed using BIT QCT software (Mindways Software, Texas, USA).

The dominant structural difference in the two groups was a lower FNaBMD in FN fracture cases compared to controls (Mean (SD) 0.49 (0.09): 0.62 (0.11) P<0.001) with no difference in area. Associated with this was an increase in inferior movement of center-of mineral mass FNaBMD (0.41 (0.10): 0.33 (0.09) P<0.001), that was correlated with FNaBMD decreases in both fracture (FNaBMD=0.53-0.1F*NaBMD) and control (FNaBMD=0.79-0.49*FNaBMD) participants. Regarding FNSigma there were no mean differences in fracture patients and controls.

In cross-sectional step-wise logistic regression analysis age, FNaBMD, FNdelta and FNSigma, but not FNW, all contributed to discrimination (FN fracture yes/no= Age*-0.05+FNaBMD*-11.67+FNSigma*-2.73+FNDelta*7.91). 

Thus the dominant anatomical factors observed to be associated with FN fractures were decreases in FNaBMD and associated increases in inferior movement of the center of mass FNDelta with a possible role for preserved cortex FNsigma.

  1. 1. Khoo BC, Lewis JR, Brown K, Prince RL. Evaluation of a simplified hip structure analysis method for the prediction of incident hip fracture events. Osteoporos Int 2016;27:241-8.