Human Amniotic Membrane Improves Healing in a Chronic,Massive Rotator Cuff Repair Model

Abstract

Objectives:

Rotator cuff tendon heals by fibrovascular scar that is weaker than native tissue leading to repairs that are prone to failure. Our objective was to investigate the ability of an amniotic membrane-derived human allograft to improve rotator cuff tendon-bone healing and skeletal muscle architecture in a chronic massive rotator cuff injury and repair model in rats. We hypothesized that application of human amniotic membrane to the tendon-bone interface at the time of repair will result in increased attachment strength and rotator cuff muscle fiber force secondary to improved bone formation and nthesis histomorphometry, less plasminogen activator inhibitor-1 (PAI-1) expression, and induced muscle repair gene expression.

Methods:

Eighty-five male Sprague-Dawley rats were allocated to one of four groups: (1) uninjured (U), (2) injury only (IO), (3) chronic injury and repair (CR), or (4) chronic injury and augmented repair (ER). The IO, CR, and ER groups underwent unilateral detachment of the supraspinatus and infraspinatus tendons with injection of the corresponding muscle with 3 u/kg Onabotulinum toxin A. Four weeks following injury, a surgical repair was performed in the CR and ER groups using transosseous suture fixation with human amniotic membrane applied to the tendon-bone repair site in the ER group. Animals were euthanized at four weeks following the repair. Biomechanical testing of the supraspinatus and infraspinatus tendon-bone complex and single fiber contractility testing of the supraspinatus muscle were performed. Microcomputed tomography was utilized to quantitate bone microstructure at the repair site. The healing tendon-bone interface was evaluated with histomorphometric quantification of fibrocartilage formation and collagen organization, and immunostaining with PAI-1. The rotator cuff muscle was evaluated with immunohistochemical localization for lipid deposition and total cross-sectional area. Real-time polymerase chain reaction evaluated relative expression of genes relevant to muscle repair. Statistical analysis was performed using a nonparametric Kruskal-Wallis test with significance set at p<0.05.

Results:

Augmentation with amniotic membrane did not lead to increased load-to-failure or stiffness compared to CR. The cross-sectional area, maximum isometric force, and specific force of individual supraspinatus muscle fibers were significantly greater with ER compared to CR (Figure 1). Microcomputed tomography revealed a larger volume of newly formed bone present at the bony footprint with ER compared to CR. Histomorphometric analysis demonstrated significantly greater fibrocartilage area and collagen organization at the healing tendon-bone insertion site with ER compared to CR at 4 week. Qualitatively, there was less PAI-1 immunostaining noted at the tendon-bone interface in the ER group. There was less lipid deposition and greater cross-sectional area of the rotator cuff muscle fibers in the ER group. Rotator cuff muscle gene expression with ER demonstrated a downregulation of apoptosis regulatory genes BCL2 and Caspase-3, adipogenesis regulatory genes FITM1 and FITM2, TNF-alpha and TGF-beta 3 compared to CR.

Conclusion:

Augmentation with human amniotic membrane in a chronic, massive rotator cuff injury and repair model led to greater bone formation, more organized tendon enthesis, induced muscle repair gene expression, and greater rotator cuff muscle fiber force secondary to less PAI-1 expression.