Primary rotator cuff repair is a common procedure that consistently yields favorable clinical results.1 Revision rotator cuff repair and reconstruction yield less consistent clinical results and are associated with a significant incidence of recurrent cuff tearing.2 Possible factors contributing to the loss of tissue continuity have included poor quality or frank loss of rotator cuff tissue, diminished biological potential of the rotator cuff tendon, and excessive mechanical stress on or strain of the reconstructive surgical construct.3
I conducted a pilot study involving a technique that addresses these potential factors, amalgamating several contemporary surgical methods with the addition of a novel step: an autogenous tendon graft incubated in autologous bone marrow concentrate.
Materials and Methods
Ten consecutive patients (7 men, 3 women) enrolled in this retrospective case series. Mean age at time of surgery was 58 years (range, 47-65 years). Mean follow-up was 24 months (range, 12-44 months), and no patients were lost to follow-up. Mean time between original primary repair and current reconstruction was 36 months (range, 6-120 months). Criteria for enrollment included unremitting shoulder pain, radiographs showing no significant degenerative joint disease, magnetic resonance imaging confirming a large (3-5 cm) full-thickness rotator cuff tear with retraction, and history of prior rotator cuff repair on the affected shoulder without associated biceps tenodesis. The intraoperative inclusion criterion was direct visualization of a 3- to 5-cm full-thickness rotator cuff tear with retraction of at least 3 cm. Validated Constant, American Shoulder and Elbow Surgeons (ASES), and University of California Los Angeles (UCLA) shoulder scoring systems were used to collect range-of-motion, pain, strength, daily function, and patient satisfaction data before and after surgery. Standard error was calculated. Two-sample t test was used for preoperative–postoperative comparisons. Postoperative integrity of the rotator cuff reconstruction was evaluated by an independent full-time academic musculoskeletal radiologist using dynamic diagnostic ultrasound (iU22 xMatrix Ultrasound System [Philips Healthcare] at L 9-3 MHz). Informed consent was obtained from each patient. The study was approved by institutional review board.
After induction of general anesthesia, each patient was placed in the lateral decubitus position. Bone marrow (60 mL) was aspirated through a 14-gauge needle from a dorsal iliac table, just inferior to the iliac crest (Figure 1). The patient was then placed into the beach-chair position on a surgical shoulder table. The aspirated marrow was centrifuged at 2800 and 3800 rpm for 14 to 17 minutes (Magellan Autologous Platelet Separator; Arteriocyte Medical Systems) to yield 10 mL of a concentrated (4- to 5-fold) mixture of platelet-rich plasma (PRP) and mesenchymal stem cells. Surgery was performed through a 3-cm oblique anterior mini-open incision between the anterior corner of the acromion and the coracoid process, as I previously described.4 The deltoid muscle was split, not detached. Acromioplasty and release of the coracoacromial ligament were performed. The rotator cuff was inspected under ×4.5 optical magnification. The cuff tissue was mobilized and débrided back to a healthy-appearing margin. The size and shape of the rotator cuff defect were then estimated. The long head of the biceps was harvested from its origin just distal to the superior glenoid labrum unto the intertubercular sulcus on the proximal humerus. The remainder of the biceps tendon was tenodesed to the surgical neck of the humerus. The biceps tendon graft was then manipulated and fashioned (by longitudinal partial-thickness incision and expansion) to fit the cuff defect (Figures 2, 3). The expanded graft was incubated in the concentrated marrow (10 mL) for 60 minutes (Figure 4). Débridement at the base of the greater tuberosity down to bleeding cancellous bone was followed by insertion of multiple bone anchors bearing several strands of No. 2 synthetic suture. These strands were then passed through the biceps tendon graft for secure fixation (Figure 5). The débrided end of the rotator cuff was then sewn to the biceps tendon graft using locking stitches under zero tissue tension with the arm in full adduction. The free end of the graft was sewn to the subscapularis tendon (Figure 6). The remaining marrow concentrate was injected both deep and superficial to the rotator cuff construct. No additional wound irrigation fluid was injected or suction drain inserted. After surgery, the patient was placed into an abduction pillow for 1 month and then engaged in passive motion for 1 month. Active-assisted motion began 3 months after surgery.
Clinically, all patients improved with respect to pain, motion, strength, function, and satisfaction by virtue of the reconstructive surgery. After surgery, mean Constant score was increased, from 13 to 71 (P < .001). Mean ASES score increased from 18 to 75 (P < .001). Mean UCLA score increased from 4 to 28 (P < .001) (Table). Ultrasound showed 0% incidence of full-thickness retearing. Dynamic scanning during abduction showed maintained reduction of the humeral head within the glenoid socket; superior subluxation of the humeral head was not detected. The biceps tendon graft was continuous with the rotator cuff tendon, indicative of graft integration: tissue healing at the graft–bone and graft–tendon interfaces (Figures 7, 8). There were no intraoperative or postoperative patient-related complications.