Intracapsular femoral neck fractures are associated with decreased cortical width and increased proportions of Haversian canals with diameters greater than the normal mean plus 3 SD (i.e., >385 μm). Such canals might be formed if closely associated resorbing osteons merge; a cortical event analogous with the loss of cancellous connectivity. To test this, we investigated the pattern of osteon distribution in the aging femoral neck to determine if remodeling osteons were distributed in anatomical clusters. Femoral neck biopsies from female patients with intracapsular hip fractures (n = 13) were compared with age/gender-matched cadaveric controls (n = 13). Solochrome-stained sections were analyzed for Haversian canal location, canal diameter, and the presence of an osteoid surface. Clustering was investigated using statistical software with a cluster defined as two or more osteoid-bearing osteon centers within 0.75 mm of each other. Clusters occurred more frequently than would be expected by chance (p < 0.001). Fracture cases had more clusters per unit area (3.14 ± 0.31 clusters/25 mm² of cortical bone) than controls (1.89 ± 0.22) (p = 0.002). In fracture cases, the antero-inferior, antero-superior, and infero-anterior regions had more clusters per 25 mm² than comparable control regions (ant/inf: 4.12 ± 0.79, 1.70 ± 0.60, p = 0.025; ant/sup: 5.31 ± 1.1, 1.80 ± 0.59, p = 0.013; inf/ant: 3.15 ± 0.49, 1.27 ± 0.29, p = 0.004). The mean number of clusters per 25 mm² per region correlated with the mean porosity per region (adjusted r² = 0.60; p = 0.014), and the total number of giant canals per region correlated with the total number of clusters per region (adjusted r² = 0.58; p = 0.011). In conclusion, remodeling osteons are clustered or grouped anatomically, and fracture cases have more clusters than controls. Our data suggest that merging of adjacent, clustered osteons during resorption could lead to the rapid development of canals with excessive diameters and focal weakness. Clustering is greatest in those regions that we have previously shown to have the largest relative reductions in bone strength compared with controls and known to be maximally loaded during a sideways fall. This implicates the remodeling process underlying clustering of remodeling osteons in the aetiology of hip fracture.
Keywords:
Clusters; Clustering; Hip fracture; Osteons; Remodeling