The effect of immobilization on rotator cuff healing using modified Mason-Allen stitches: a biomechanical study in sheep.

Long-term biomechanical integrity of the surgical site is crucial to the outcome of rotator cuff repair. Usually after rotator-cuff surgery, the arm is immobilized for 6-12 weeks to minimize disruption at the healing site. Unfortunately, currently there is very little clinical data on the effects of immobilization after rotator cuff surgery. This study analyzed the effect of immobilization on rotator cuff healing by comparing the healing strength of the tendon in a bone trough using a sheep model. In sixteen (16) female adult sheep, the right infraspinatus tendons were reattached into a bone trough using a modified Mason-Allen suture pattern. After surgery, each sheep was randomly assigned into either an immobilized or non-immobilized treatment group. The immobilized group was fitted with a softball taped under the foot to restrict limb movement for 6 weeks. The non-immobilized treatment group was allowed to fully bear weight on the treated shoulder. At 26 weeks, the animals were sacrificed and the surgical shoulders were harvested and stored in a -30 degrees C freezer. Biomechanical testing was performed using a MTS 809 Axial/Torsional Servohydraulic System. Special cryo-grips were used to test the infraspinatus tendon-bone interface in anatomical alignment. All samples were preloaded with 10 +/- 5 N, and then force was measured during a crosshead displacement of 500 mm/min until failure. Force and displacement were acquired at 100 Hz and used to calculate load-to-failure and stiffness of the bone-tendon interface. The load-to-failure (N) and stiffness (N/cm) were determined for both treatment groups. For the non-immobilized group, the loads to failure and stiffness means were 2571.95 +/- 329.9 N and 2319.09 +/- 457.72 N/cm. The immobilized group means were 2954.42 +/- 473.52 N (load-to-failure) and 2579.13 +/- 383.51 N/cm (stiffness). A two-tailed two-samples unequal variance student t-test with significance set at 0.05, found no significant difference between the treatment groups for load-to-failure (p = 0.2174) and stiffness (p = 0.8286).