A Scoping Review of Biomechanical Testing for Proximal Humerus Fracture Implants

David Cruickshank; Kelly A. Lefaivre; Herman Johal; Norma J. MacIntyre; Sheila A. Sprague; Taryn Scott; Pierre Guy; Peter A. Cripton; Michael McKee; Mohit Bhandari; Gerard P. Slobogean


BMC Musculoskelet Disord. 2015;16(175) 

In This Article


The literature describing the biomechanical testing of proximal humerus fracture implants is broad and heterogeneous. It is evident that biomechanical testing is being performed more frequently to compare proximal humerus fracture treatments; however, significant limitations to the clinical utility of the current testing models exist. These include a lack of models for three- and four-part fractures and a high variability in the testing parameters utilized.

The most common model identified was the simulated two-part fracture. From a practical perspective, this is not surprising since the fracture (osteotomy) occurs in the surgical neck region and does not require the investigator to recreate fractured tuberosity fragments or impaction of the humeral head. Two-part fractures are also appealing to model because fixation is easily achieved in the humeral head and shaft, and mechanical testing can focus on axial, bending, and torsional loads across a single fracture line. Despite the study design advantages of focusing on two-part fractures, it is our opinion that three- and four-part fractures represent the true surgical challenge and should be the focus of most biomechanical testing.[7,8] Fourteen studies simulated a three-part fracture, and only five studies used a four-part model.

Another key finding of our scoping review was the substantial heterogeneity in testing parameters. We found almost no duplication of testing configurations and minimal standardization, which would allow comparison between studies. Consequently, we classified the studies based on biomechanical testing themes such as direction of force and testing mode. In most studies, the direction of force could be placed into one of three categories: torque, axial load, or cantilever bending (varus or valgus). In addition to variations in the direction of force applied, a wide range of force magnitude and cycles were observed. For example, 33 studies used cyclic loading to test their constructs; however, the number of loading cycles used in each study ranged from 5 to 1,000,000 cycles. Furthermore, in many of these studies the magnitude of the force applied was not reported, or there was a wide variety in combinations of forces.

Similar heterogeneity was also observed in the reporting of cadaveric specimens used. Authors commonly did not report the age of the specimens or the pre-testing analysis conducted to ensure the validity of results. Only 57 % of studies reported the age of the specimens and 67 % reported their pre-testing analysis. Specifically, fewer than half of the studies reported the bone mineral density of their specimens, which is essential for ensuring testing specimens are comparable and the results can be interpreted within the context of other published studies.

The final gap identified in our scoping review was the lack of biomechanical testing of arthroplasty implants in proximal humerus fracture models. Although there are likely many studies that test the mechanical properties of shoulder arthroplasty implants in an intact humerus, only five studies were identified that performed testing within a PHF model. This lack of relevant testing is important to recognize because the implantation of a humeral arthroplasty stem in the setting of a proximal humerus fracture is technically challenging and inherently unstable due the displacement of the tuberosity fragments. Furthermore, given the exponential increase in reverse total shoulder arthroplasty for PHF patients, relevant biomechanical testing would provide invaluable information to help guide treatment decisions.[2]