Bone resorption of the greater tuberosity after open reduction and internal fixation of complex proximal humeral fractures: fragment characteristics and intraoperative risk factors

Hypothesis and background: In complex proximal humeral fractures, bone resorption of the greater tuberosity is sometimes observed after open reduction and internal fixation (ORIF). However, this has not been well characterized, and risk factors for resorption are not completely understood. We aimed (1) to identify the risk factors associated with bone resorption of the greater tuberosity and (2) to quantify the geometric and bone density characteristics associated with bone resorption using 3-dimensional computed tomography models in complex proximal humeral fractures treated with ORIF.

Methods: We identified a retrospective cohort of 136 patients who underwent ORIF of 3- or 4-part proximal humeral fractures; greater tuberosity resorption developed after ORIF in 30 of these patients. We collected demographic, fracture-related, and surgery-related characteristics and performed multivariable logistic regression analysis to identify factors independently associated with the development of greater tuberosity resorption. Furthermore, we identified 30 age- and sex-matched patients by use of propensity score matching to perform quantitative fragment-specific analysis using 3-dimensional computed tomography models. After the fragment of the greater tuberosity was identified, the number of fragments, the relative fragment volume to the humeral head, and the relative bone density to the coracoid process were calculated. Measurements were compared between matched case-control groups.

Results: We found that an unreduced greater tuberosity (odds ratio [OR], 10.9; P < .001), inadequate medial support at the calcar (OR, 15.0; P < .001), and the use of an intramedullary fibular strut (OR, 4.5; P = .018) were independently associated with a higher risk of bone resorption. Quantitative fragment-specific analysis showed that greater tuberosities with a larger number of fragments (5 ± 2 vs. 3 ± 2, P = .021), smaller fragments (9.9% ± 3.8% vs. 18.6% ± 4.7%, P < .001), and fragments with a lower bone density (66.4% ± 14.3% vs. 88.0% ± 18.4%, P = .001) had higher rates of resorption.

Discussion and conclusion: An unreduced greater tuberosity or inadequate medial support increases the risk of greater tuberosity resorption, as do a larger number of fracture fragments, smaller fragments, and lower bone density. Additionally, fibular strut grafting is an independent risk factor for tuberosity resorption. Further study is needed, but alternatives to strut grafting such as femoral head allograft may warrant serious consideration.