PURPOSE:
To compare the biological bone-to-tendon healing using three different medialized bone bed preparation techniques (i.e. cortical bone exposure, cancellous bone exposure, and no cartilage removal) in a rat model of medialized rotator cuff repair.

METHODS:
Twenty-one male Sprague-Dawley rats with 42 shoulders were subjected to bilateral supraspinatus tenotomy from the greater tuberosity. The rotator cuff was repaired using medialized anchoring with the cortical bone exposed, the cancellous bone exposed, or no cartilage removed. Four and three rats in each group were killed for biomechanical testing and histological evaluation, respectively, at postoperative 6 weeks.

RESULTS:
All rats survived until the end of the study, but one infected shoulder in the cancellous bone exposure group was excluded from further analysis. Compared with the cortical bone exposure and no cartilage removal groups, the rotator cuff healing of the cancellous bone exposure group showed significantly lower maximum load (cancellous bone exposure group: 26.2 ± 2.3 N, cortical bone exposure group: 37.6 ± 7.9 N, no cartilage removal group: 34.6 ± 7.2 N, P = 0.005 and 0.029) and less stiffness (cancellous bone exposure group: 10.5 ± 2.4 N/mm, cortical bone exposure group: 17.4 ± 6.7 N, no cartilage removal group: 16.0 ± 3.9 N, P = 0.015 and 0.050) at postoperative 6 weeks. In all three groups, the repaired supraspinatus tendon healed towards the original insertion rather than the medialized insertion. The cancellous bone exposure group showed inferior fibrocartilage formation and insertion healing.

CONCLUSIONS:
The medialized bone-to-tendon repair strategy does not guarantee complete histological healing, and the removal of excessive bony structure impairs bone-to-tendon healing. This study concludes that surgeons should not expose the cancellous bone during the medialized rotator cuff repair.