Acellular Dermal Matrix in Rotator Cuff Surgery
The success of rotator cuff repair (RCR) surgery can be measured clinically (validated outcome scores, range of motion) as well as structurally (re-tear rates using imaging studies). Regardless of repair type or technique, most studies have shown that patients do well clinically. However, multiple studies have also shown that structurally, the failure rate can be very high. A variety of factors, including poor tendon quality, age over 63 years, smoking, advanced fatty infiltration into the muscle, and the inability of the tendon to heal to bone, have been implicated as the cause of the high re-tear rate in RCRs. The suture-tendon interface is felt to be the weakest link in the RCR construct, and suture pullout through the tendon is believed to be the most common method of failure. This review of the published literature seeks to determine if there is support for augmentation of RCR with acellular dermal matrices to strengthen the suture-tendon interface and reduce the re-tear rate.
Rotator cuff repairs (RCRs) can be challenging due to poor tendon quality and the inability of tendon to heal to bone. Smoking, age over 63 years, fatty infiltration, and massive cuff tears are all factors implicated in increased failure rates.1-3 Tears >3 cm have a structural failure rate ranging from 11% to 95% in the literature.1-5 Massive tears (tears >5 cm or involving 2 or more tendons) are even more complex and have failure rates of 20% to 90%.5,6 The weakest link in the RCR construct is the suture-tendon interface, and suture pullout through the tendon is thought to be the most common method of failure.6 The purpose of this review is to examine whether literature supports the use of acellular dermal matrices (ADMs) in rotator cuff surgery.
The high rate of structural failures after RCR has led surgeons to seek means to augment repairs and new means of reconstruction for irreparable tears. Freeze dried allograft tendons have been used historically with mixed results, including reports of complete graft failures and foreign body reaction.7-10 Porcine intestinal submucosal membrane “patches” gained popularity due to off-the- shelf availability of the graft. However, these were found to have poor outcomes with early graft rejection and intense inflammatory reaction.11,12 Recently, ADMs have gained significant interest due to favorable biomechanical properties and clinical outcomes.13-19
An ADM is an allograft composed of mostly type I collagen that is processed to remove donor cells while preserving the extracellular matrix. There are several commercially available ADMs with different methods of processing and sterilization, as well as handling characteristics.20,21 In vivo studies have demonstrated that removing the cellular components allows infiltration of native cellular agents, such as fibroblasts, vascular tissue, and tenocytes, while causing minimal host inflammatory reaction.21-23 In addition, superior suture pullout strength has been demonstrated by multiple benchtop and preclinical studies.23,24 Therefore, ADMs play a dual role of strengthening the repair while allowing infiltration of host cells and growth factors to potentially promote healing at the repair site.
Emerging Evidence
Multiple biomechanical studies have evaluated ADMs in RC models.24-28 Barber and colleagues24 demonstrated that ADM had significantly higher loads to failure (229 N) than porcine skin (128 N), bovine skin (76 N), and porcine small intestine submucosa (32 N) (P < .001). In another study, Barber and colleagues25 subsequently demonstrated, in a cadaver RC tear model, an increase in mean failure strength in augmented repairs with ADM (325 N) compared to cadaveric controls (273 N) (P = .047).
A subsequent study by Barber and Aziz-Jacobo26 compared ADMs to a control model of allograft RC. The ADMs had significantly higher tensile modulus (P < .001) and higher suture retention measure by a single-pull destructive test of a simple vertical stitch (P < .05) than the RC allograft. The ultimate load to failure of the ADM model was higher than the RC allograft control (523±154 N vs 208±115 N); however, this difference did not reach statistical significance.26 Beitzel and colleagues27 evaluated ADM augmentation in a cadaver RC model and found a statistically significant increase in load to failure in ADM augmented repairs vs nonaugmented controls, (575.8 N vs 348.9 N, P = .025). Ely and colleagues28 also demonstrated that repairs augmented with ADM had a higher load to failure (643 N vs 551 N) and less gap formation (2.2 mm vs 2.8 mm) compared to controls, although this difference was not statistically significant.
