A Morphological Study of Age-Related Changes in Medullary Characteristics of Proximal Humerus
Proximal humeral fractures (PHFs) are a significant clinical concern, particularly in patients over the age of 60 who often suffer from severe osteoporosis. While many PHFs can be managed non-operatively, complex fractures necessitate surgical intervention. In older individuals, the loss of trabecular structure and the calcar region in the proximal humeral medullary cavity predisposes them to comminuted fractures and loss of medial hinge support, leading to high rates of implant failure and reoperation. Recent advancements suggest that endosteal augments incorporated into locking plate constructs may offer improved medullary support and mechanical stability. However, the variability in the shapes, modes, and positions of these augments can result in implant failure and reduction loss, underscoring the need for a deeper understanding of the morphological changes in the proximal humerus across different age groups.
This study aimed to analyze and compare the anatomical degeneration patterns of the medullary canal between older and younger patients to enhance our understanding of the medullary morphology of the proximal humerus and its clinical implications for endosteal support. The research was conducted retrospectively, reviewing patients with PHFs treated at a single trauma center from January 2013 to December 2016. Inclusion criteria required that the contralateral proximal humerus was unaffected and that patients were at least 18 years old. Exclusion criteria included previous operative treatment on the contralateral upper limb, a history of illnesses affecting the shape and function of the contralateral upper limb, long-term use of steroids or other drugs affecting bone mineral density (BMD), and combined metabolic osteopathy.
Patients were divided into two groups based on age: a younger group (<60 years) and an older group (≥60 years). All included patients underwent a full-length computed tomography (CT) scan of the uninjured humerus. Using the Mimics software (version 18.0; Materialise Inc., Leuven, Belgium), three-dimensional (3D) models of the uninjured humerus were reconstructed from the CT images in the Digital Imaging and Communication in Medicine format. The reconstruction process involved creating masks for the whole humerus, cortical bone, medullary canal, and cancellous bone based on specific Hounsfield unit (HU) thresholds. The humerus models were then imported into the 3-matic software (V11.0, Materialise, Belgium) for detailed analysis.
Measurement benchmarks were established for the medullary canal in both groups. The medullary canal in the region 20 to 60 mm below the humeral head is approximately cylindrical, and the axis of the medullary cavity of the humeral shaft (humerus canal axis [HCA]) was generated using the axis fitting tool of 3-matic. The humeral head was fitted to a sphere using the sphere fitting tool, with the center of the sphere designated as the center of the humeral head (Os) and the semidiameter of the sphere as Rs. The coronal datum plane (CP0) was established based on HCA and Os, and the axial datum plane (AP0) was established with HCA as the normal line, crossing the lowest point of the humeral head on CP0. The intersection point of HCA and AP0 was designated as O. On the view of CP0, the most medial and lateral intersection points of the medullary canal and AP0 were marked as medial point (M) and lateral point (L), respectively. The highest point of the medullary cavity model was defined as the apex point (V). The interface between the medullary cavity and the humeral head and its subcapital region was defined as the support plane (SP), established through the points of V and M and perpendicular to CP0. The supporting angle (SA) was defined as the angle between AP0 and SP. The section of the medullary canal at the level of the SP was calibrated, and the best fitting circle of the section was obtained using the automatic fitting arc function of 3-matic software, with the diameter (Ds) of the circle measured.
Statistical analysis was performed using SPSS (version 21.0; SPSS Inc., Chicago, IL, USA). The Kolmogorov-Smirnov test was used to assess the normality of the metric data. The independent sample t-test was used to compare the differences of metric data between groups conforming to the normal distribution, reported as arithmetic mean ± standard deviation. Count data were tested using the Chi-square test. Multiple regression analysis was used for outcomes of interest, such as proximal volume (PV), proximal height (PH), proximal medial distance (PM), and proximal lateral distance (PL), adjusting for gender and humerus radius. Statistical significance was set at P < 0.050.
Sixty patients were included in the study. Demographic characteristics showed no significant differences in gender, dominant side, height, and weight between the younger and older groups. However, there were significant differences in age, BMD, and Neer classification of the injured side. The values of Rs were 22.31 ± 1.72 mm and 22.06 ± 2.06 mm in the younger and older groups, respectively, with no significant difference. The PV in the older group increased significantly (21.60 ± 3.03 cm³ vs. 29.40 ± 1.98 cm³, P < 0.001), with significant expansion in the proximal (PH: 25.36 ± 1.08 mm vs. 27.38 ± 0.67 mm, P < 0.001), medial (PM: 13.14 ± 1.16 mm vs. 21.60 ± 3.03 mm, P < 0.001), and lateral (PL: 17.90 ± 0.95 mm vs. 26.02 ± 0.76 mm, P < 0.001) directions. The SA was also decreased in older patients (65.28° ± 9.89° vs. 45.02° ± 2.10°, P < 0.001), but did not show significant variation among the older patients (ranging from 41.04° to 49.24°). Multiple regression analysis results were consistent with the unadjusted analysis, showing a significant difference in medullary cavity morphology. Additionally, there was cancellous bone in the humeral head in the older group, with its junction with the medullary cavity forming a quasi-circular facial structure with a diameter of 11.67 ± 1.50 mm.
Previous studies on the morphology of the proximal humeral medullary cavity are limited. Sprecher et al. reported that in osteoporotic patients, the loss of trabecular bone in the greater tubercle and the medial metaphysis was more significant compared to the loss in the subchondral region. However, these studies were based on the analysis of the two-dimensional plane of cadaveric bone. Through 3D reconstruction of the medullary structure, this study revealed that the cylindrical-shaped medullary cavity in younger patients becomes irregularly “bellbottom” shaped with bone loss in older patients. This morphological change may explain why single implants fail to provide effective medullary support in older patients. Furthermore, older patients exhibited a significant medial offset of the medullary cavity, a characteristic change that should be considered when placing endosteal support as close to the medial cortex and humeral head as possible.
The age-related differences in the morphology of the medullary cavity are associated with the clinical prognosis of PHF. Clinical studies have shown that younger patients have better clinical outcomes with nails- or plate-based medullary augmentation compared to older patients. This is because existing nails or augments are mostly columnar structures, which are well-matched with the medullary cavity of younger patients. In contrast, the expansion of the proximal humeral medullary cavity in older patients renders the cavity irregular, making it anatomically incompatible with the shape of existing nails or struts. Therefore, more effective anatomical supporting augments are needed for the treatment of older patients.
In conclusion, compared to younger patients, the medullary canal of the proximal humerus expands significantly in older patients, particularly in the medial offset, resulting in an “eggshell”-like cavity structure. The residual dense cancellous bone in the humeral head could serve as an endosteal support structure with a relatively fixed support angle. These findings provide anatomical references for the development of reasonable endosteal augmentation strategies for older patients with PHF.
doi.org/10.1097/CM9.0000000000001597
Was this helpful?
0 / 0