Various attempts have been made to reattach the torn tendon edge to the anatomical insertion site. However, if a tear is too large to reattach the tendon to the anatomical site, the torn edges are fixed to the bony trough created medial to the original insertion site. It is clinically important to know how far medially we can shift the reattachment site. Ten fresh frozen cadaveric shoulders (mean, sixty- eight years old) were used. The specimen was mounted to a custom-made testing device with a 22-N joint compressive load. Medial shift was simulated by the placement of suture anchors along the lines 3,10, and 17 mm medial to the greater tuberosity. The ranges of glenohumeral motion (abduction, external rotation at 0 and 60° of abduction, and internal rotation at 0 and 60° of abduction) were measured before and after medial shift. The range of motion was measured by goniometer under constant torque application. Medial shift (3,10, and 17 mm) reduced the range of abduction, compared with that of the intact shoulder, by 4 ± 5°,17 ± 10°, and 31 ± 110, external rotation by 8 ± 5°,11 ± 6°, and 18 ± 7°, external rotation at 60° of abduction by 8 ± 5°,15 ± 10°, and 18 ± 8°, internal rotation by 4 ± 4°,7 ± 7°, and 11 ± 7°, and internal rotation at 60° of abduction by 5 ± 5°,7 ± 5°, and 9 ± 7°. Joint motion decreased significantly in almost all directions after a medial shift of 10 mm and 17 mm. Significant restriction of joint motion occurs after a cuff repair when a bony trough is created more than 10 mm medial from the foot print.
