The incidence of rotator cuff tears increases with age, thus the rotator cuff tear is often associated with osteoporotic or osteopenic bone in the proximal humerus, especially with female patients. For testing of fixation devices such as suture anchors used in rotator cuff repair often animal bones are used. They are easily to obtain, inexpensive and some have been found to be similar to human bone. But can we rely on the results drawn from these studies in our daily surgical practice?

The purpose of this study was to compare the trabecular bone mineral density, the trabecular bone volume fraction and the cortical layer thicknes in the greater tubercle in different species to evaluate their infiuence on primary stability of suture anchors under a cyclic loading protocol representing the physiologic forces placed on rotator cuff repairs in vivo. We hypothezised that maximum pullout forces as well as the modes of failure are different for a suture anchors in different humeri. The available three different types of anchor fixation design (screw: Spiralok 5mm, Super Revo 5mm, press-fit: Bioknotless RC, wedging: Ultrasorb) were tested. The bone mineral density (BMD) of the humeri was measured by a 64-slice-computed tomography system. Each anchor was tested individually until failure. The sutures were pulled at 135° to the axis of the humeral shaft, simulating the physiological pull of the supraspinatus tendon. Starting with 75 N the tensile load was gradually increased by 25 N after everey 50 cycles until failure of the anchor fixation system occurred. The ultimate failure load, the system displacement after the first pull with 75 N and the mode of failure were recorded.

The ultimate failure loads of each anchor were different in the human osteopenic, human healthy, ovine and bovine humeri. The statistical significancies for pull out forces between the anchors varied from species to species. The biomechanical testing of suture anchors for arthroscopic rotator cuff repair in ovine and bovine humeri does not give reliable data that can be transferred to the human situation. The significances between the suture anchors found in ovine and bovine humeri are different from the results in human humeri. When taking the impaired bone quality of older patients into account the results from ovine and bovine humeri are even less predictable. We found a positive correlation between maximum failure load and cortical layer thickness for the Super Revo and the Ultrasorb anchor. The ultimate failure load seems to depend mainly on the cortical thickness and on the subcortical trabecular bone quality.