INTRODUCTION
Rotator cuff tears are a common cause of shoulder pain and dysfunction, with rotator cuff repair being the primary surgical intervention to restore tendon integrity and shoulder function [1]. Despite advances in surgical techniques, retear rates remain unacceptably high, ranging from 20% to 75% depending on the tear size and patient factors [2-5]. This high retear rate has prompted ongoing research into adjunctive therapies and techniques to enhance the biomechanical and biological properties of the repair site [6-10]. One such approach is the use of patch augmentation to provide additional mechanical support and serve as a scaffold for biological healing [6,11].
Patch augmentation is intended to offload mechanical stress from the tendon–bone interface, thereby reducing the risk of early repair failure [12]. Additionally, the patch becomes incorporated into the repaired tendon, providing further reinforcement to the tissue. Numerous patch materials, including synthetic meshes [13], acellular dermal matrices (ADMs) [14], and collagen-based patches [15], have been evaluated for their potential to strengthen repairs and promote tendon healing. In particular, patches composed of ADM have been reported to offer clinical advantages due to their favorable healing profile, structural integrity, and biocompatibility [16]. However, the immediate time-zero mechanical benefits of these patches remain unclear. Although several studies have demonstrated improved tendon–bone integration and decreased retear rates with patch augmentation [11], it is uncertain whether those benefits arise from enhanced mechanical stability at the time of surgery or the promotion of biological healing over time. To address that gap, studies using controlled biomechanical models to specifically evaluate the acute mechanical reinforcement provided by ADM patches at time zero are needed. Prior research has largely focused on biological integration, so further investigation into the immediate load-sharing capacity of these patches is essential to improve understanding of their potential contribution to early repair strength.
Another important consideration is the impact of surgical techniques on the effectiveness of patch augmentation. Factors such as suture configuration, knot tying, and tensioning can influence the biomechanical performance of a repair. In particular, no one has thoroughly investigated whether the presence or absence of knot tying at the medial row affects the distribution of forces across the repair site and the overall stability of the construct [17-19]. Understanding such nuances is critical because optimizing the technique for patch incorporation could lead to better outcomes, both biomechanically and biologically.
Our primary aim in this study was to use porcine tendons to evaluate the biomechanical impact of ADM patch augmentation on the initial stability of rotator cuff repair and investigate whether medial knot tying influences the mechanical properties of the construct. Our hypothesis was that patch augmentation would enhance time-zero stability by providing greater resistance to elongation and a higher load to failure, compared with non-augmented repairs, and that medial knot tying would further improve stability by increasing compression at the tendon–suture interface.
METHODS
This study used porcine flexor digitorum profundus (FDP) tendons to model rotator cuff repair and evaluate the biomechanical effects of patch augmentation. FDP tendons were selected based on their size and availability, which ensured a reliable and reproducible model for time-zero biomechanical testing. To minimize variability, all tendons were obtained from the hind legs of pigs freshly slaughtered at a local abattoir and prepared in a standardized manner. Because this study was conducted at time zero, the use of tissues from already slaughtered pigs was considered more ethically appropriate than sacrificing live rabbits or rats. In this way, ethical concerns were minimized, and institutional review board approval was deemed unnecessary. This study used porcine specimens obtained from a local butcher after standard slaughter for food consumption. Therefore, ethical approval from the Institutional Animal Care and Use Committee was not required.
Tendon Harvest Techniques
Upon arrival, each hind limb was thoroughly rinsed with saline solution to remove contaminants. The limbs were then placed in a prone position on a dissection table, with the plantar surface facing upwards. A longitudinal incision was made along the posterior aspect of the hind limb, starting from the plantar surface and extending proximally toward the mid-thigh. The skin and subcutaneous tissue were carefully retracted to expose the underlying muscular structures. Care was taken to preserve the surrounding tissues during this step to prevent damage to the tendons. The deep digital flexor muscle was located posterior to the tibia and fibula. The FDP tendon was identified by palpating the muscle belly and tracing the tendon distally toward its insertion at the phalanges. The tendon was distinguished from other nearby structures by its position deep within the leg and its continuation to the digits in multiple slips. The FDP tendon was carefully freed from surrounding connective tissue using blunt dissection. Proximal and distal attachments were identified and severed with a scalpel, ensuring that a sufficient length of tendon was retained for the subsequent experiments. The tendon was detached distally at the level of its insertion to the phalanges and proximally at the musculotendinous junction.
Once it had been excised, each tendon was immediately rinsed with saline solution to remove any residual blood or debris. The tendons were then stored in sterile containers with saline-moistened gauze and kept at 4 °C until further processing. The harvested tendons were inspected for any signs of macroscopic damage or contamination. Only tendons free from defects were included in the study. The detailed dimensions of each tendon are listed in Table 1. According to the literature, the average thickness of the human supraspinatus tendon is 4.9±2.1 mm [20]. In our study, the porcine specimens exhibited tendon thicknesses comparable to these human values. We preserved the maximum possible tendon length during specimen preparation to ensure adequate purchase by the biomechanical testing apparatus.
Repair Techniques
All repairs were performed using a standardized double-row suture bridge technique, with medial and lateral row anchors placed at predetermined intervals to ensure uniformity. Each tendon was secured to the porcine calcaneus bone analog using a combination of medial (Healmax, 4.8 mm double-loaded anchors; Osteonic Co., Ltd.) and lateral row anchors (Biocombi, 4.75 mm, Osteonic Co., Ltd.) placed at 1-cm intervals. The reversing half-hitch alternating post (RHAP) knot involves a five-throw configuration, with alternating posts and reversing half-hitches to maintain tension and stability throughout the knot-tying process (Fig. 1). The sequence starts with two half-hitches tied in the same direction, followed by three alternating throws that reverse the direction and switch the standing post, distributing forces evenly across the knot [21]. The 18 FDP tendons used in this study were divided into three groups, with six tendons in each group. The groups were divided based on the augmentation and knotting techniques (Fig. 2, Table 2).
Group 1: Non-patched Group (Non-patched+Tendon Knotted)
Group 1 (control group) tendons were repaired to the porcine calcaneus without any augmentation. The medial row anchors were placed, and each suture limb was passed through the tendon. After the sutures were passed through the tendon, they were tied using the standard RHAP knot-tying technique, and then the limbs were brought laterally and secured to the lateral row anchors to complete the double-row construct.
Group 2: Patched without Medial-Row Knotting (Patched)
In group 2 (patch-augmented group without knot tying), a 20 mm×15 mm ADM patch (BellaCell HD, HansBiomed Corp.) of 3 mm thickness was used for augmentation, and no medial row knots were tied. After the medial row anchors were placed and the sutures were passed through the tendon, the patch was applied over the tendon without securing the sutures with a knot. The suture limbs were then passed through the patch and secured to the lateral row anchors.
Group 3: Patched with Medial-Row Knotting (Patched+Tendon Knotted+Patch Knotted)
In group 3 (patch-augmented group with knot tying), the patch augmentation was performed with medial knot tying. The sutures from the medial row anchors were first passed through the tendon and tied using three standard square knots. The patch was then placed over the tendon, and the suture limbs were passed through the patch and tied using an additional three square knots before they were secured to the lateral row anchors.
Biomechanical Analysis
Strain measurement of the tendon
The length change of the tendon was measured using a Mercury RT strain measurement system (Mercury Systems) that consists of a video extensometer and Mercury RT Lite software. The software offers sub-pixel accuracy in tracking surface patterns to enhance the reliability and sharpness of the measurement results. The standard Mercury RT system achieves a strain resolution of up to 10 microstrain (10–5) units. Elongation was defined as the change in the length of the tendon–suture construct under tensile load. The two-dimensional locations of 5 landmarks (p1 to p5) along the tendon were automatically tracked by the Mercury system. Because the bottom landmark (p1) was set as a fixed point, it was placed laterally to maintain its initial position even when it was subjected to tension (Fig. 3). The vertical length (l1) of the 1st segment (A1), connecting p1 and p2, was calculated with a trigonal equation, l1=A1∙sin (90-θ), where θ is the angle between the 1st segment and a vertical line. After the mechanical test, the angle θ on the video clips was measured using ImageJ v1.54h (U. S. National Institutes of Health). As the angle θ decreased when subjected to tension, l1 increased. The other segmental lengths (l2 and l3) were considered to remain vertical during tension. The vertical length between p4 and p5 was excluded from monitoring because in some cases, p5 went outside the camera’s visual area as the tendon was stretched. The strain was calculated as the ratio of the elongation to the initial length of the tendon.
Tensile tests
Tensile tests were performed using a StepLab EA005 electro-mechanical testing system (STEP Engineering S.R.L.). The specimen was mounted to the mechanical tester’s upper jig in a reproducible manner. The initial angle between the elongated tendon and loading direction was about 20° oblique to the vertical so that the initial direction of the tendon’s elongation was perpendicular to the lateral flat bone surface where the lateral anchors were inserted (Fig. 4). The tendons were mounted securely, and tension was applied by moving the upper loading head up at a rate of 0.5 mm/sec until ultimate failure. Load to failure was defined as the maximum load sustained by the construct before complete failure. Stiffness was calculated as the slope of the load-elongation curve, reflecting the construct’s resistance to deformation.
Statistics
Statistical analyses were performed using analysis of variance (ANOVA) with post-hoc comparisons between groups to evaluate the significance of differences in biomechanical outcomes. Prior to ANOVA, tests for normality (Shapiro-Wilk test) and homogeneity of variances (Levene’s test) were conducted to ensure the appropriateness of a parametric analysis. Post-hoc multiple comparisons were conducted using the Bonferroni correction method. All statistical tests were conducted in IBM SPSS statistics version 22.0.0 (IBM Corp.), and a P-value of less than 0.05 was considered statistically significant for all comparisons.
RESULTS
The experimental setup consisted of three groups, with six specimens prepared for each group. However, due to improper fixation during testing, only five datasets were successfully obtained for each group. Among them, one experiment from group 2 and one from group 3 exhibited premature failure at the medial anchor that occurred before tendon failure. Those two datasets were excluded from the final analysis. As a result, the data analysis was performed on five datasets from group 1, four datasets from group 2, and four datasets from group 3. The primary outcome measures were the elongation, strain, change in angle, load to failure, and stiffness, and each was evaluated at time zero following repair.
Strain of the Tendon
The l1 segment bore the most tendon elongation in all three groups (Table 3). The elongations of segments l2 and l3 were kept around 1 mm. In the elongation of l1, the mean elongation was 5.0±5.3 mm in group 1, 12.9±6.5 mm in group 2, and 4.4±4.4 mm in group 3. Group 2 exhibited significantly greater elongation than both group 1 (P=0.016), and group 3 (P=0.027). However, group 1 and group 3 did not differ significantly from each other (P=0.47). The mean strain of l1 was 35.2%±38.9% in group 1, 75.8%±42.9% in group 2, and 29.1%±28.9% in group 3. The difference in the strain of l1 was significant (P=0.028) between group 2 and group 1 and showed a trend toward statistical significance between group 2 and group 3 (P=0.052). However, group 1 and group 3 did not differ significantly from each other (P=0.374). As shown in Fig. 5, the elongation and strain of l1 indicate that tendon patch augmentation did not have a large effect on the elongation of the repair construct, but knotting the medial tendon to the bone limited the elongation of the near-bridge tendon.
Failure Load of the Repair
Data for 13 of the initially prepared 18 specimens were analyzed as results (Table 4). Five specimens were excluded from these results because they tore at the interface (P4 or P5 in Fig. 3) with the upper metal loading head of the tensile testing machine. The 13 specimens analyzed experienced tendon failure around P2, where the tendons were fixed by medial anchors or/and sutures. The load-to-failure measurements were similarly comparable among groups. The mean load to failure was 447.0±166.1 N in group 1, 337.5±135.3 N in group 2, and 339.3±53.9 N in group 3 (Table 4). The groups did not differ significantly in the failure load (P>0.5). The absence of significant differences indicates that patch augmentation did not enhance the ultimate strength of the repair construct at time zero.
DISCUSSION
The findings of this study indicate that patch augmentation using ADMs does not improve biomechanical stability at time zero compared with non-augmented repairs, regardless of the use of medial knot tying. These results are unexpected given the theoretical advantages of patch augmentation, such as improved load distribution and internal splinting of the repaired tendon site [22]. ADMs, which are treated to remove cellular components and thereby reduce the possibility of triggering an immune response but retain the collagen-based extracellular matrix, offer both biological and mechanical advantages in tendon healing [23]. They provide a robust scaffold that supports the regeneration of new host tissue and contribute strength to the repaired area [23,24]. Biologically, they promote cellular proliferation, tissue integration, and vascularization, enhancing the natural healing process [24]. Mechanically, ADMs reinforce the tendon–suture interface, providing structural support that improves load-bearing capacity and reduces elongation during early repair [25]. Beitzel et al. [26] used cadavers to investigate the biomechanical effects of various scaffold types on the initial strength of rotator cuff repairs and reported that scaffold augmentation, specifically using a collagen patch or human dermis patch, significantly increased the ultimate load to failure, compared with non-augmented repairs.
Numerous studies have reported that patch augmentation during rotator cuff repair can significantly reduce retear rates. For example, Andrade et al. conducted a systematic review demonstrating that the use of patch augmentation led to a notable reduction in retear rates and pain scores, while also enhancing shoulder functionality [6]. The incorporation of a patch into the tendon is thought to enhance the mechanical stability of the repair site and promote tendon healing through increased tissue organization and vascularization [6]. On the other hand, a recent systematic review of randomized controlled trials of rotator cuff repair with patch augmentation reported that limited evidence supports the association of patch augmentation with improved patient-reported outcomes [11].
Although the outcomes of patch augmentation are somewhat debated, comprehensive studies investigating the biomechanical effects of the patch at time zero—the period immediately following surgery before the patch is fully incorporated into the tendon—are lacking. One study used nine matched pairs of cadaveric shoulders to compare single-row rotator cuff repair with and without ADM augmentation [27]. The augmented group, in which ADM was sutured medially to the tendon and fixed laterally to the humerus, showed a significantly higher maximum load to failure (560.2±95.5 N vs. 345.7±60.8 N), though the linear stiffness was lower (65.2±15.6 N/mm vs. 77.2±15.7 N/mm), likely due to the graft’s elasticity. In this study, we hypothesized that patch augmentation would enhance time-zero stability by providing greater resistance to elongation and a higher load to failure than non-augmented repairs. However, contrary to our hypothesis, patch augmentation did not positively influence any biomechanical outcome. Table 5 summarizes biomechanical studies on ADM patch augmentation in rotator cuff repair [26,28]. One potential explanation for the lack of improvement in time-zero mechanical strength with ADM patch augmentation is the interface between the patch and the native tendon. At the time of implantation, the patch is not biologically integrated and thus might not contribute significantly to the mechanical continuity of the repair construct. The ADM functions primarily as a scaffold for tissue ingrowth and remodeling, which occur gradually over time. Without integration, the patch might behave more like a passive overlay than an active load-sharing component. Thus, although the patch might have beneficial effects once it becomes incorporated into the underlying tendon at the repair site over time [6,8,11], it has no significant effect when it is simply placed on top without integration. Additionally, the mechanical contribution of the ADM could be limited by its compliance relative to the native tendon tissue, resulting in insufficient tension-sharing during initial loading. Our findings suggest that the mechanical advantages of ADM patches might not manifest until biological incorporation occurs, highlighting the importance of long-term studies that evaluate both immediate and delayed structural contributions.
Another key finding of this study is that the load to failure of the repair was not significantly affected by whether medial row knots were tied before patch application, but tying the knot on the medial tendon to bone limited the elongation of the near-bridge tendon. Few studies have compared rotator cuff repair with and without medial row tying during patch augmentation. However, recent studies on rotator cuff repairs without patch augmentation suggest that even without tying the medial row, the biomechanical properties achieved are comparable to those obtained with tying [19,29]. Mijares et al. [29] compared the biomechanical outcomes of double-row rotator cuff repair with and without medial row knot tying, focusing on displacement, stiffness, and ultimate load to failure. Their results showed no significant differences between the knotted and knotless techniques in any of the measured parameters, suggesting that knotless medial row repair can provide biomechanical stability comparable to knotted repairs. Other studies have raised concerns that medial row knot tying could increase the risk of retear at the musculotendinous junction [17,18]. In this study, on the other hand, the absence of knot tying on the medial tendon resulted in reduced mechanical stability, leading to greater elongation and strain that indicated insufficient tendon fixation. These findings highlight the critical role of medial tendon tying in enhancing fixation strength and preventing excessive elongation and strain, thereby contributing to improved biomechanical stability. However, due to the lack of literature comparing medial knot tying with non-tying during patch augmentation, there could be limitations in applying previous studies comparing medial knot tying and non-tying in double-row suture bridge rotator cuff repairs when discussing the results of this study. Therefore, caution should be exercised when interpreting these findings in the light of previous research.
One limitation of this study is that it was performed in ex vivo conditions, which do not fully replicate the complex in vivo environment of rotator cuff repair. Factors such as joint motion and muscle forces and biological processes such as inflammation and tissue remodeling all contribute to the ultimate success of a repair. Thus, although time-zero biomechanical testing provides valuable insights into the immediate stability of different constructs, it cannot predict the results when the tendon is integrated into the overall structure. In vivo studies are needed to evaluate the effects of patch augmentation throughout the course of tendon healing, comparing time-zero with time after healing, with a focus on both biomechanical and histological outcomes. In addition, the small sample size in each group might have introduced potential bias, such as reduced statistical power and increased variability, which could have contributed to variations in the results and limited the generalizability of our findings. Another limitation is that only 13 of the originally intended 18 specimens were used for the experiment. Additionally, although elongation, strain, change in attachment angle, load to failure, and stiffness were measured, the pressure applied to the repaired cuff tendon was not assessed. We found no differences in the parameters we measured, but analyzing the contact pressure could reveal whether the patch increased pressure at the repair site. Such a finding could support previous studies showing improved healing outcomes with patch augmentation. Lastly, it is important to note that the porcine FDP might not accurately mimic the bone characteristics of the human proximal humerus, which could lead to differing results when applied to human subjects.
CONCLUSIONS
ADM patch augmentation did not enhance time-zero mechanical strength under the tested conditions, though biomechanical advantages might emerge after integration to the tendon. In addition, patch augmentation without medial row knot tying showed the highest elongation and strain among the groups, potentially reducing the stability of the repair.
PDF Links
PubReader
ePub Link
XML Download
Full text via DOI
Full text via PMC
Download Citation
Print
