The torsional strength of cemented implants is likely influenced by stem geometry. Five straight stems with different cross-sectional shapes (circular, oval, triangular, round-rectangular, sharp-rectangular) were custom-machined. The stems were cemented into tubes using bone cement, and subjected to torsion (2.5deg/min)(n=7). At initial failure (crack through the cement mantle or loss of cement-stem adhesion), the sharp-rectangular stem resisted over 33% more torque than the other four stems (p=0.13). At ultimate failure (5° stem rotation), the resistance provided by the circular stem was less than 12% of either rectangular stem (p< 0.05). Additional studies are needed to determine the effects of long-term loading.

To determine the influence of cross-sectional implant stem shape on cement failure under torsional loading.

The sharp-rectangular stem provided the greatest torsional resistance against initial failure. At ultimate failure, the two rectangular stems performed better than the other stems, with the circular stem providing the least torsional resistance.

A stem design that provides increased resistance to torsion will, in all likelihood, improve the longevity of cemented stemmed implants.

Five straight stems with different cross-sectional shapes (circular, oval, triangular, round-rectangular, sharp-rectangular) were custom-machined. Each stem was cemented in a square aluminum tube using Simplex® cement (Stryker, Michigan, USA). A materials testing machine was used to apply torsion to the stem at 2.5 deg/minute until failure. ‘Initial failure’ was defined as the appearance of a crack through the cement mantle and/or the loss of cement-stem adhesion. ‘Ultimate failure’ was defined as 5° of stem rotation. Results (n=7) were compared using one-way ANOVAs with post-hoc Student-Newman-Keuls tests (p=0.05). The sharp-rectangular stem withstood over 33% more torque at initial failure than the other stems (p=0.13). At ultimate failure, the circular stem provided significantly less torsional resistance than the other four stems (p< 0.05), and was able to resist less than 12% of the torque applied to either rectangular stem. These results suggest shape may play a role in the onset of implant loosening due to torsion. Further studies are required to explore other shapes and to examine the effects of cyclic loading and cement soaking.

Funding: Natural Sciences and Engineering Research Council, University of Western Ontario

Purpose: To compare the torsional stability provided by five implant stems with different cross-sectional geometries under cyclic loading.

Methods: Cemented stems with five different cross-sectional shapes – circular, oval, triangular, rectangular with rounded edges (round rectangular), and rectangular with sharp edges (sharp rectangular) – were custom machined from stainless steel. Stem dimensions were selected to fit within the humeral canal (based on a 6mm x 8mm dimensioning scheme) and shapes were based on commercially available-designs. Seven specimens of each stem shape were tested. ||The stems were potted in square aluminum tubes using bone cement, and allowed to cure for 24 hours prior to testing. A materials testing machine and a custom designed loading fixture were used to apply torsion to the stems. A sine wave loading pattern was applied until ultimate failure (5° of stem rotation) was reached. This loading pattern had a lower bound of 0.9Nm and an upper bound that started at 4.5Nm and was increased in increments of 2.25Nm every 1500 cycles. The load was cycled at 2Hz. Statistical analyses on both the number of cycles and torque to failure were performed using one-way ANOVAs followed by post-hoc Student-Newman-Keuls (SNK) tests (p< 0.05).

Results: Overall, ANOVAs showed an effect of shape on the number of cycles (p< 0.0001) and torque to failure (p< 0.001). SNK tests revealed the sharp rectangular stem provided the greatest resistance to torque (p-cycles< 0.001; p-torque< 0.001) compared to all other stems. Other significant differences resulted in the following ranking of the shapes: sharp rectangular, round rectangular, triangular, and circular = oval.

Conclusions: The results of this study agree with static testing previously conducted on the same set of stem shapes. Although the sizes of the stems were chosen to roughly replicate upper limb implants, these results may be extrapolated to larger stems such as for the hip or knee. To improve implant longevity, it is important that the best fixation possible be obtained through all available avenues, including improved cementing techniques, and optimal implant designs. An alteration in implant stem shape may assist in achieving this goal.

Three 3mm transverse slices were sectioned from the distal cancellous region of seven fresh-frozen cadaveric humerii. Each slice was marked with a 3x3mm grid, and subjected to compressive testing using a flat cylindrical indenter (1.6mm diameter). Indentation modulus and strength were calculated for each site, and pooled into nine anatomically-defined regions. The most distal slice had higher moduli values (p< 0.05), and the posterior capitellar region had lower moduli values (p< 0.05). There were no slice or regional differences in strength. This suggests that surgical procedures requiring cancellous fixation utilize the most distal aspect of the humerus while avoiding the posterior capitellum.

To quantify the indentation strength and modulus of distal humeral cancellous bone, and identify any regional variations.

Cancellous bone modulus in the distal humerus decreases from distal to proximal. The posterior capitellum has a lower modulus than the other regions of the distal humerus.

The influence of slice depth emphasizes the importance of minimizing the amount of bone removed during prosthetic replacement. Regional variations in modulus suggest that the posterior capitellum should be avoided during fixation of implants or placement of screws.

Three 3mm transverse cancellous bone slices obtained from the distal end of each of seven fresh-frozen cadaveric specimens were subjected to compressive testing using a materials testing machine with a 1.6mm flat cylindrical indenter. Testing was performed in a 3x3mm grid. The indentation modulus and local strength were calculated for each test site, and then averaged into nine regions defined by the capitellum, medial and lateral trochlea, and anterior, central and posterior sections for each slice. Mean modulus was found to be 309.8±242.0 MPa (range: 2.9–1041.7 MPa). Yield strength averaged 4.4±2.5 MPa (range: 0.6–16.3 MPa). The highest modulus was found in the distal-most slice (p< 0.05). The lowest modulus region was the posterior capitellum (p< 0.05). There were no differences in strength between slices or across the nine regions. A comparison with proximal tibial cancellous bone properties suggests the distal humerus may carry loads approaching 30% of those at the knee, assuming that bone adapts to stress magnitudes.

Funding: Natural Sciences and Engineering Research Council; University of Western Ontario