Development of a Rabbit Chronic-Like Rotator Cuff Injury Model for Study of Fibrosis and Muscular Fatty Degeneration

A reproducible and physiologically-relevant animal model provides the ability to advance the understanding of disease pathogenesis, evaluate the outcomes of clinical therapies, and improve and further develop surgical treatments³⁵. In this study, a reliable and accurate rabbit SSC model that mimics aspects of human RC anatomy and pathophysiology was established. RC tears are related to progressive and likely irreversible muscular degenerative changes, resulting in a reduced healing potential. For example, Ko et al. showed that the reattachment of rabbit SSP at 6 weeks did not reverse muscle atrophy or FD in the following 6 weeks. Such FD-mediated muscle atrophy influences several important clinical parameters, including tendon-muscle strength and joint range of motion, which may affect the surgical outcomes^36,37.

The protocol established here showed significant chronic-like attributes after the transection of SSC muscle-tendon units. Specifically, these changes include visibly decreased muscle mass and increased adipose content and fibrotic tissue (Figure 2, Figure 3, and Figure 4). These findings are consistent with degenerative changes reported in human RC tears³⁸. In recent years, the rat has emerged as one of the most intensively studied animal models for RC disease and injury due to its high anatomical similarities with both human and rat SSPs traveling under the acromion^38,39,40. However, it should be noted that the portion of rat SSP which passes under the acromial arch is muscular as opposed to tendinous, which is the case in humans⁴¹. Most importantly, Barton et al. recognized a lack of significant fat accumulation after SSP tendon detachment in rats²³, which stands in contrast to the human condition⁴². As such, it is believed that the rabbit SSC complex may provide an appropriate model to mimic the chronic RC tear of humans.

To ensure the reproducibility of this model, two points are worth noting when performing this protocol. First, after the transection of muscle-tendon units, the free-end of the transected tendon may be at risk of forming adhesions, which can make tendon retrieval challenging for subsequent manipulations. To avoid this issue, a non-resorbable silicone tubing was used to wrap the free end of the muscle-tendon junction following transection to avoid spontaneous adhesion to surrounding tissues as well as spontaneous healing (Figure 1E). Further, the transected muscle-tendon unit during a second procedure for intervention (i.e., to perform a secure repair; data not shown) can be clearly identified by wrapping the end of injured tissues at the time of initial surgery. This technique is economical, effective, and can be easily implemented in surgery⁴³. Second, rabbits are a highly sensitive species that may exhibit detrimental behavior following surgery. To avoid such issues, it is highly recommended that a soft collar is also applied to prevent undesired behavior, including self-mutilation, licking of surgical sites, and removal of sutures (Figure 1I). Compared to commercially conventional E-collars that are made of rigid plastic, the self-made soft collar did not cause any skin injury or other side effects that affected animal welfare or the quality of scientific inquiry. Together, such steps are critical to create an accurately reproducible rabbit RC injury model and provide the possibility for studying the regenerative repair strategies.

To study tendon pathophysiology and healing in an animal model, a distinct and reproducible injury must be created, and the study time points must be carefully selected. The vast majority of studies on tendon injury and healing have been performed on fully transected animal tendons⁴⁴, as transection is a simple procedure that is highly reproducible and can adequately simulate the clinical scenario^45,46. Huegel et al. showed that the injury of a partially transected tendon was less severe than that of a fully transected tendon, and immobilization had a detrimental effect on tendon mechanics, including increased joint stiffness⁴⁷. To evaluate the atrophy and FD that is seen in the setting of massive RC tear, it is essential to define the experimentally observed characteristic time points. Gupta et al. have validated a RC injury model in the male rabbit and observed muscle atrophy at 2 and 6 week time points, with increased fat content at later time points (less than 5% fat content at 2 weeks vs. more than 10% fat content at 6 weeks), consistent with the pathological process observed in human RC tears¹¹. In this study, a massive RC tear was created by transection of the SSC muscle-tendon unit in male and female rabbits for 4 weeks, which resulted in SSC muscle FD (36.5% fat content). Thus, a 4 week time point is appropriate for generating SSC muscle FD in male and female New Zealand white rabbits.

Several limitations to this study exist. These include: (i) steps associated with animal model generation, such as a relatively short time point and potentially inflammatory materials (silicone-based penrose tubing) for chronic-like injury generation; (ii) animal model characterization and analysis, such as lack of gait analysis and electromyography to assess joint kinematics and muscle contractile force generation; and (iii) animal model comparison, such as lack of comparison with other RC injury sites.

In terms of model generation, human RC injuries typically involve progressive atrophy and FD that may occur over the span of several years, which is relatively longer than the 4 week time point reported here. This is deemed to be acceptable, since an animal model that generates around 36.5% intramuscular fat in a relatively short time span will be logistically convenient and can be prolonged if deemed necessary. Moreover, the biocompatibility of silicone-based implants, such as penrose tubing, has been a source of long-standing controversy due to reports of cellular immune response and inflammation⁴⁷; therefore, an alternative inert material, such as polyethylene glycol (PEG), may be substituted for wrapping the resected tendon if pursuing inflammation-associated RC studies.

In terms of animal model characterization and analysis, the lack of gait analysis⁴⁹ and electromyograph studies⁵⁰ may limit the study's findings to qualitative histological data. These aspects may be addressed in future studies by using video motion analysis⁵¹ and surface electromyography⁵⁰ to generate quantitative data on shoulder kinematics and RC muscle performance.

In terms of model comparison, since the SSP and infraspinatus tendons in the rabbits have also been widely used for RC studies, comparing injury severity, including FD among these different injury sites in the future, will identify additional sites for model optimization.

In summary, this study has developed a protocol for modeling chronic-like RC injuries in male and female rabbits. This model is convenient for investigators owing to its simplicity (transection) and relatively short period to induce chronicity (4 weeks) while generating a large degree (36.5%) of intramuscular FD. As such, this protocol is expected to aid investigators in the study of RC pathophysiology, as well as facilitate the development of novel therapeutics for muscle-tendon repair and regeneration.