INTRODUCTION:​ Replacement of the diseased shoulder joint with implants has been complicated by a high incidence o f glenoid component loosening. A better understanding of the mechanical properties of glenoid cancellous bone coupled with measurements of implant stability would permit the development of a more reliable method of glenoid component fixation. This study assessed the stiffness of cancellous bone throughout the glenoid cavity, correlated these with quantitative computer tomography (CT) measurements and made an assessment of implant stability for some basic fixation modes..

METHOD :​ Eight normal adult cadaveric glenoids were scanned in the sagittal and transverse planes. Three millimeter thick slices were cut at 3, 6, and 9 mm below the articular surface with a diamond blade. Mechanical indentation testing was performed with an Instron testing machine, using a 3 mm diameter flat indentor.

To assess implant stability, the articulating surface of an additional nine fresh frozen cadaveric glenoids were prepared following the established protocol for the Neer II component. Four iterations of keel design were evaluated:​ no keel, half keel, full keel and cemented keel. Linear Variable Displacement Transducer's were utilized to measure micromotion of the component. The implants were subjected to three block randomized loads to test the effect of load magnitude, direction and angle.

RESULTS:​ Regional variations in bone quality:​ No significant difference was found between elastic stiffness measured as indentation (modulus - K) of the first and second slices. There was, however, a decrease at the 6-9 mm post-articular level. Superior regions were found to be the strongest within each slice, followed by anterior regions, while infero-posterior trends varied between slices. A least squares linear regression between volume fraction for each location revealed a wide variation in the correlation. When all data were pooled, there was a similar lack of correlation between CT density measurements and mechanical stiffness.

Component Stability:​ The cemented component demonstrated significantly less distraction than other uncemented fixation mechanisms. Generally, there were no significant differences in component stability among uncemented keel designs. Further, significant differences were noted for both uncemcnted and cemented keel designs at the inferior site when the eccentricity c f the superiorly applied load was changed from 22.5° to 30°.

DISCUSSION AND CONCLUSIONS:​ The lack of a consistent correlation between mechanical and radiological data may be due to the anisotropic nature of cancellous bone. As mechanical strength is dependent upon the bone's modulus and trabecular orientation, the use of quantitative CT alone may not be an adequate predictor of bone strength. Examination of regional variation indicate that superior sites are more appropriate for mechanical fixation. The large reduction in bone properties below 6 mm indicates that longer glenoid component anchoring mechanisms may not significantly contribute to component fixation. The universal lack of any significant difference among uncemented keel designs suggests that a keel may not improve stability relative to an unkeeled polyethylene component. However, it is suspected that if keel length was increased to make contact with the cortical shell, the component would generate sufficient reaction forces to resist micromotion and reduce the incidence of glenoid component loosening.