Recent advances in shoulder research

Shoulder pathology is a growing concern for the aging population, athletes, and laborers. Shoulder osteoarthritis and rotator cuff disease represent the two most common disorders of the shoulder leading to pain, disability, and degeneration. While research in cartilage regeneration has not yet been translated clinically, the field of shoulder arthroplasty has advanced to the point that joint replacement is an excellent and viable option for a number of pathologic conditions in the shoulder. Rotator cuff disease has been a significant focus of research activity in recent years, as clinicians face the challenge of poor tendon healing and irreversible changes associated with rotator cuff arthropathy. Future treatment modalities involving biologics and tissue engineering hold further promise to improve outcomes for patients suffering from shoulder pathologies.


Introduction
As the elderly population expands, so do age-related orthopaedic disorders. Th is is of particular concern in the fi eld of shoulder pathology, as both osteoarthritis (OA) and rotator cuff disease are degenerative conditions that increase in the aging population. Th ese represent the most common causes of pain and disability, and have been subjects of research and treatment innovation in recent years. Arthritis of the shoulder can have a number of etiologies. Osteoarthritis, trauma, avascular necrosis, infection, and infl ammatory arthropathies can all lead to loss of cartilage integrity and destruction of the joint surfaces. Loss of cartilage and incongruent joint surfaces result in painful articulation, necessitating orthopaedic treatment. Osteoarthritis is the most common cause of shoulder arthropathy and has been linked to age [1] and chronic overuse [2]. Conventional treatment options include nonsteroidal anti-infl ammatory medication, cortisone injections, arthroscopic debridement, and joint replacement. It is expected that the rate of upper extremity arthroplasty will soon double and lead to increased health care costs and societal burdens [3].
Rotator cuff disease occurs in an age-related fashion and can exist along a spectrum, from rotator cuff tendinitis, to partial thickness rotator cuff tears, to full thickness rotator cuff tears. A recent study by Yamamoto and colleagues [4] demonstrated that the prevalence of cuff tears in a Japanese village was 20.7% for its general population; the risk factors identifi ed for tears were history of trauma, arm dominance, and age. Others have shown that, in asymptomatic shoulders, an increased prevalence of tears is associated with increased age [5,6]. Tears can enlarge with time, and the increase in tear size is associated with retraction of the muscle tendon unit, which can lead to changes in muscle architecture [7], joint mal-alignment, and altered biomechanics [2]. Not all tears are symptomatic, although rotator cuff repair of painful tears is one of the most common orthopedic procedures in the US. In this review, we discuss the two major degenerative disorders of the shoulder, OA and rotator cuff disease, as well as new insights into how to treat these debilitating conditions.

Glenohumeral osteoarthritis Etiology and pathology
OA, also known as degenerative joint disease, is defi ned as non-infl ammatory degeneration of the cartilage and narrowing of the glenohumeral joint space. Radiographic fi ndings of glenohumeral OA include joint space narrowing, circumferential osteophyte formation, subchondral cyst formation, posterior wear or bone loss of the glenoid, and/or subchondral sclerosis. Arthritis of the shoulder has many etiologies, including primary (idiopathic) and secondary (post-traumatic or developmental). Primary gleno humeral OA is considered rare [8], yet it is becoming an increasingly recognized source of pain and disability in the shoulder. Overuse and trauma have been correlated with increased risk of developing degenerative joint disease [9]. Increased intrinsic glenoid retroversion has been shown to lead to increased wear of the posterior glenoid, and these individuals are prone to earlier onset OA [10]. Chronic overuse of the shoulder can lead to excessive wear of the articulating surfaces with eventual thinning of the articular cartilage [11].
Post-capsulorraphy arthropathy is arthritis associated with surgical procedures for the treatment of instability, particularly from over-tightening of the anterior soft tissues of the shoulder. Over-tightening of the soft tissues leads to reduced external rotation and increased compressive load on the posterior articular cartilage of the glenoid, resulting in accelerated development of arthritis in the shoulder. Mal-positioned and migrated hardware, such as suture anchors or loosened screws, may also cause mechanical damage to the glenohumeral joint cartilage [12]. More recently, thermal injuries from soft tissue shrinkage devices used during shoulder arthroscopy have led to chondrolysis, and eventual OA [2].
Th e radiographic fi ndings described above are in contrast to arthritis associated with infl ammation (e.g., rheumatoid arthritis). Rheumatoid arthritis in the shoulder typically presents with medial wear of the glenoid, absence of osteophyte formation, large cyst formation, and osteopenia.

Current treatment options for osteoarthritis
Th e management of shoulder OA typically begins with nonoperative modalities before surgery is considered. Th e mainstays of non-operative treatment include oral and injected analgesics and anti-infl ammatory drugs, physical therapy, and lifestyle modifi cations [13]. Th e goal of physiotherapy is to increase joint range of motion and strengthen muscles of the scapular girdle. Arthritic joints are prone to stiff ness, and the stiff ness is often a source of pain. Joint infi ltration of local analgesics, often combined with steroidal anti-infl ammatory drugs, is common and is often performed in out-patient settings. Steroid injections and visco-supplementation may provide short-term pain relief and help physicians diagnose intraarticular pathologies [14,15]. Such treatments are not without risk, however, and may lead to chronic degenerative changes to the joint and attenuation of the soft tissues in and around the joint [16].
When nonoperative treatment options no longer alleviate symptoms and symptoms interfere with daily activities and sleep, surgical management is often considered ( Table 1). Treatment options include arthroscopic debridement, cartilage repair, and biological and arthroplastic replace ment [17]. Arthroscopic debridement, with or without capsular release, may provide short-term relief of pain in the osteoarthritic shoulder; however, deterioration over time can be expected for most patients due to the loss of cartilage thickness and the inability to regenerate lost tissue. Some studies have demonstrated the benefi t of this procedure as an early temporal bridge to arthroplasty [18,19].
A major limitation to the consideration of an arthroplasty in a young patient is the longevity of the prosthesis. Th e rate of survival of a shoulder arthroplasty over the long term (15 to 20 years) is approximately 85% [20]. For young patients, alternatives are considered in order to avoid future revisions. Biologic resurfacing with a soft tissue interposition with or without humeral head replace ment has led to controversial results [18,[21][22][23]. Resurfacing with knee meniscus, Achilles allograft, anterior shoulder capsule, and other materials have all been reported. Early reports of biologic resurfacing were favorable, but recent mid-term results have indicated a high rate of failure and subsequent revision. Currently, glenoid resurfacing with biologic interposition is only recommended in young patients, in their third or fourth decade. Humeral head prosthetic resurfacing with stemless implants has also been considered in younger patients for humeral lesions [22]. Th e rationale is that a smaller resurfacing implant preserves proximal humeral bone stock, in the interest of future revision surgery. While this procedure has the potential advantage of minimal bone loss without humeral canal reaming, it is specifi c to treating small focal lesions or isolated humeral head arthrosis and may have little application in the setting of more severe OA [24][25][26].
Total shoulder arthroplasty (TSA) is the gold standard treatment for severe glenohumeral OA [27]. Th e growth rate for TSA continues to rise compared to other orthopedic joint replacement surgery rates [3]. Approximately 45,000 patients in the US undergo total shoulder arthroplasty or hemiarthroplasty each year [3]. A total shoulder arthroplasty involves replacement of the humeral head and prosthetic resurfacing of the glenoid ( Figure 1). A hemiarthroplasty refers to humeral head replacement alone.
Total shoulder arthroplasty off ers reliable pain relief, predictable improvement of function, and improved quality of life for a variety of shoulder arthropathies, including primary OA [28]. Th e main concern with TSA is the potential for loosening of the glenoid component over time, as this represents the most common complication [29,30]. Fixation [31] and material composition [32] are also factors related to the success of prostheses, as diff erent materials and implantation methods may infl uence osteolysis and risk of arthroplasty revision. Recent results, however, support the longevity of polyethy lene glenoid resurfacing [23]. Improvements in glenoid component materials and engineering are an ongoing subject of research.
Hemiarthroplasties are used mostly for select cases of arthritis, such as OA in a younger individual and rheumatoid arthritis in which bone loss precludes implan tation of a glenoid component and rotator cuff arthropathy [28,33]. In general, hemiarthroplasty demonstrates inferior results when compared to TSA [27,[34][35][36]; however, such procedures may be more appealing in certain settings, such as for very young patients, patients with severe bone loss, and patients with avascular necrosis involving only the humeral head [37].
A reverse shoulder arthroplasty is arguably one of the most important contributions to the treatment of certain shoulder arthropathies in the past several years. Reverse total shoulder arthroplasty is indicated primarily in the setting of rotator cuff insuffi ciency and rotator cuff arthropathy, but has also been used for fracture treatment, revision of failed shoulder arthroplasty, and sequelae of trauma. Reverse TSA prostheses have a ballin-socket design, with a semi-circumference ball being implanted in the glenoid and a stem with a concave polyethylene cap implanted in the humerus ( Figure 1). Rotator cuff arthropathy represents a spectrum of shoulder pathology characterized by rotator cuff insuffi ciency, diminished acromiohumeral distance, and arthritic changes of the glenohumeral joint [38]. Reversal of the  [28,39].

Rotator cuff disease Etiology and pathology
Rupture of one or more of the rotator cuff tendons from the humeral head is one of the most common orthopedic injuries in the US, with over 250,000 repairs performed each year [3]. Large rotator cuff tears, which include more than one of the rotator cuff tendons, lead to increased morbidity and probability of post-surgical repair-site failure [40]. Additionally, many factors may perpetuate the likelihood of a failed repair, including age, gender, severity and duration of injury [41,42]. Regardless, the predictability of failure and the factors associated with impaired healing and reduced strength of the repaired rotator cuff are currently unclear. In both the clinic and in animal models, changes in muscle architecture and structure have been associated with chronic rotator cuff disease [7,43,44], and a role for rotator cuff muscle health and rehabilitation in healing repairs has been suggested [45,46]. Following a rotator cuff tear, fatty accumulation [47] and atrophy [44] are thought to play a role in the reparability of the tendon-to-bone insertion, as these factors can lead to increased repair site tension due to tenomuscular retraction and muscle fi brosis [48]. Recent animal studies have elucidated the pathomechanisms of fatty degeneration of the rotator cuff muscles after chronic cuff tears [49][50][51]. Rotator cuff tears in rodents led to accumulation of adipocytes, intramuscular fat globules, and intramyocellular fat droplets in the injured muscles ( Figure 2). Adipogenic and myogenic transcription factors and markers were upregulated in the injured rotator cuff muscles, and the severity of changes was associated with tear size and concomitant nerve injury. Th e status of the rotator cuff by serial ultrasound examination after large and massive rotator cuff repairs has been used to potentially delineate failure mechanisms as having either mechanical or biological causes. High tension at the repair site immediately post-surgery may increase the risk for mechanical failure of the repair site [48]. Failure of the repair site may also result from a lack of appro priate healing at the tendon-to-bone insertion [52]. Improvements in both tendon-to-bone insertion strength as well as promotion of tendon-to-bone healing are precedent in order to advance the success of rotator cuff repair strategies.

Current treatment options for rotator cuff disease
Treatment modalities for rotator cuff disease are dependent on the severity of degeneration and symptoms of the patient; various surgical treatment options are outlined in Table 1. While acute, traumatic rotator cuff tears can be treated surgically with high success rates and marginal morbidity, treatment of chronic rotator cuff disease is less promising. Surgical repairs of chronic rotator cuff tears are less likely to heal than acute repairs, and 30 to 94% of arthroscopic repairs of large, chronic rotator cuff injuries have the potential to fail, particularly within the fi rst 2 years [40,53]. Even with current repair techniques, including arthroscopic double-row repairs, failure rates post-repair remain high [54,55], suggesting  the potential role of aging and degeneration in recurring tendon-to-bone failure [56][57][58][59]. Th e healing tendon-tobone insertion following rotator cuff repair is dissimilar to the native insertion, which demonstrates four distinct transitional zones: bone, calcifi ed fi bro cartilage, uncalcifi ed fi brocartilage, and tendon. Instead, the healing tendon forms a fi brovascular scar tissue and is biomechanically weaker and more prone to failure than the native insertion [60]. Even though patients symptomatically improve after surgically repaired rotator cuff tears regardless of the structural status of the cuff in the postoperative period, studies suggest a better clinical outcome when the repair remains intact and the cuff heals back on the greater tuberosity [40,61]. Moreover, some investigations show that increased age, larger tear size, poorer muscle quality, delamination of the tendons, and longer follow-up are all related to lower healing rates and inferior clinical results [41,42]. Miller and colleagues [62] recently investigated the potential mechanisms of failed repairs by verifying the chronological status of the rotator cuff using serial ultrasound examination after rotator cuff repairs. Th e great majority of the recurrent tears (seven of nine) occurred very early in the postoperative period, perhaps suggesting a mechanical cause for the failure. Nevertheless, it remains unclear whether mechanical or biological reasons are responsible for impaired cuff healing and failure. Eff orts have been made in order to improve the initial fi xation strength and to better recreate the normal anatomical footprint of the rotator cuff . Th e use of the double-row technique, or the addition of another row of suture anchors to the fi xation construct to improve structural function and re-establish the rotator cuff footprint, has been implemented in recent years [42]. While these techniques may lead to higher healing rates, multiple clinical studies have not shown a translation from improved insertion strength to better functional or clinical outcomes [63,64]. It is likely that, even with the Normal muscles showed no fat. After tenotomy of the supraspinatus (SS) and infraspinatus (IS) tendons, the infraspinatus muscle had more intramuscular fat than the supraspinatus muscle. The 16-week specimens had more intramuscular fat than the 8-week specimens within each group. Note that grading was semi-quantitative in nature; statistical comparison and error bars were therefore not calculated. Reproduced with permission from [50].

Future therapies and treatment modalities
Th e development of therapies to improve the healing rate and functional outcomes after the onset of OA and rotator cuff damage is currently driven by several factors, including functional improvements in the strength of the repair, targeted biochemical signaling of the repair site to encourage the healing process, and establishment of native cartilage and/or tendon-to-bone insertion. Tissue engineering may provide avenues for encouraging growth, healing, and remodeling of injured musculoskeletal tissues, particularly of articular cartilage lining the glenohumeral joint and the tendon-to-bone insertion of the rotator cuff ( Figure 3).

Scaff olds and grafts
While surgery remains the last resort option for treatment of severe shoulder degeneration, compelling reasons exist to prolong the need for arthroplasty solu tions, especially in younger patients. Bioengineered devices, such as tissue engineered grafts, are currently being developed and may play a substantial role in the healing and structural maintenance of glenohumeral articular cartilage [23,65,66]. Most approaches incor porate one or more components of the tissue engineering paradigm des cribed in Figure 3. Recently, Gobezie and colleagues [66] implemented an all-arthroscopic total shoulder cartilage resurfacing technique for treatment of advanced gleno humeral OA in young patients. Using osteoarticular allo grafts from cadaveric tibial plateaus and humeri, bipolar cartilage resurfacing demonstrated early success and re habili tation [66]. Similarly, Krishnan and colleagues [23] performed successful glenoid resurfacing with fascia lata autograft in some patients and Achilles tendon allograft in others. In this study, the use of allogenic resurfacing material showed promise for reducing postoperative pain as well as for excluding donor site morbidity that is observed when using autografts [23]. Glenoid resurfacing has proven successful after 3 to 6 years following implementation of a xenograft patch seeded with pluri potent cells [65]. However, due to progressive gleno humeral space narrowing, the durability of biologic soft tissue interposition grafts may present a long-term concern [21]. Additionally, disease transmission and host rejection are potential issues when implementing bio logically derived materials from allogenic or xenogenic sources. Nonetheless, chondrocyte and osteochondral plugs and articular grafts from autogenic or allogenic sources have successfully demonstrated their potential as biologic alternatives to debridement for the treatment of cartilage defects [13,[67][68][69], and more research is needed to better interpret their effi cacy.

Rotator cuff repair Scaff olds and grafts
A number of scaff olds have been used clinically in an eff ort to augment rotator cuff tendon-to-bone repair. However, a review of currently available scaff olds by Derwin and colleagues revealed that further work is necessary to optimize scaff old properties [70]. Clinically available scaff olds lack an appropriate recreation of the native tissue's gradation in properties between the compliant tendon and the stiff bone. To address this lack of complexity, laboratory tissue engineering work has focused on a number of approaches. Biphasic [71] and triphasic scaff olds [72] have been generated and seeded with multiple cell types. Th ese studies demonstrated the importance of signaling between the various tendon-tobone cell types for generation of a functional insertion. More recent approaches have also attempted to create continuous gradients in composition and properties in order to recreate the interface that is seen at the natural rotator cuff tendon-to-bone insertion. To this end, electrospun polymer nanofi ber scaff olds were synthesized with gradations in mineral, mimicking the mineral gradation seen at the native insertion [73]. Th e gradation in mineral content resulted in a spatial variation in the stiff ness of the scaff old. Similar results were reported using a cell-seeded collagen scaff old with a gradient in retrovirus encoding an osteogenic transcription factor [74]. A tissue engineered scaff old with a gradation in properties and seeded with the appropriate cells and biofactors may ultimately provide a solution to the clinical problem of tendon-to-bone healing.

Biological aids
Th roughout soft tissue healing, several growth factors and catabolic molecules have been shown to regulate scar formation and remodeling [75]. Such alterations in biomarker production can provide insight into the normal biological response of the healing tendon, cartilage, and bone. Th e use and/or combination of exogenous growth factors, stem cells, and bioengineered scaff olds may demonstrate potential in encouraging healing and repair of the rotator cuff [76][77][78][79][80]. Th e use of individual molecules, such as bone morphogenetic protein-2 (BMP-2) and transforming growth factor (TGF)-β3, to aid cartilage and tendon-to-bone healing has been explored [78][79][80][81][82][83][84][85], yet it is likely more benefi cial to incorporate a cocktail of growth factors to best promote healing, of which the constituents are currently unknown [78]. Regardless, harmonious signaling initiated by both anabolic and catabolic factors during healing is what will likely drive the most successful repair to minimize scar formation and encourage the redevelopment of organized glenohumeral cartilage and tendon-to-bone insertion [78].
A targeted approach for enhanced repair using single growth factors embedded in scaff olds has been attempted in animal models in an eff ort to enhance rotator cuff [75] and cartilage repair [86]. Two recent studies demonstrated that TGF-β3 may accelerate healing [80,87]. Th is growth factor has been implicated in fetal development and scarless fetal healing and, thus, addition of TGF-β3 at the repaired tendon-to-bone insertion may enhance healing. Manning and colleagues [87] used a scaff old with controlled release of TGF-β3 to encourage tendon-to-bone healing in a rat rotator cuff repair model. TGF-β3 treatment led to increases in infl ammation, cellularity, vascularity, and cell proliferation in the early period after surgical repair. Th e growth factor also promoted improvements in mechanical properties compared to controls. Cellular and gene transfer approaches have shown promise for improving rotator cuff repair as well. Gulotta and colleagues [88] delivered mesenchymal stem cells (MSCs) to the rotator cuff repair site in rats, but did not see improvements in healing. Positive results were only seen after MSCs were transfected with scleraxis (Scx), a transcription factor that is necessary for tendon development [79]. Rotator cuff repairs that received Scx-transfected MSCs had higher strength and stiff ness compared with the non-transfected MSC repairs. In a similar study, MSCs transfected with membrane type 1 matrix metalloproteinase (MT1-MMP), a factor that is upregulated during embryogenesis at tendon-bone insertion sites [84], showed signifi cant improve ments in healing compared to controls. Increased fi brocartilage production was noted at the repair site along with improvements in mechanical properties. Although both targeted growth factors and MSCs show great promise for enhancing rotator cuff repair, further safety and effi cacy studies are needed to determine if results from animal studies can be applied eff ectively in the human surgical setting.
Recently, great interest has been given to biologic augmentation with platelet-rich-plasma (PRP). PRP is a solution of concentrated platelets prepared from autologous blood that contains numerous growth factors, including platelet-derived growth factor (PDGF), vascular endothelial growth factor (VEGF), and TGF-β1s [89]. Such factors make PRP an attractive option for the enhancement of recruitment, proliferation, and diff erentiation of cells in the repair site of soft tissue damage. Th e creation of higher-quality tissue at the repair site would likely enhance healing rates and clinical outcomes [90]. However, recent studies have demonstrated confl icting reports on the effi cacy of exogenous supplementation of PRP for improving healing rates and improving clinical and functional outcomes [91,92]. In rotator cuff repair augmentation, Castricini and colleagues [91] recently investigated the use of PRP on patients with small or medium tears. In this study of patients with small (<1 cm) and medium (1 to 3 cm) rotator cuff tears, augmentation with PRP at the time of rotator cuff repair did not improve Constant scores, tendon footprint thickness, or tendon thickness compared to repairs that were not augmented [91]. On the other hand, Randelli and colleagues [92] recently showed an accelerated improvement in clinical scores 3 months after surgery for patients treated with PRP at rotator cuff repair compared to those not treated with PRP. Patients treated with PRP also demonstrated reduced pain scores at 3, 7, 14, and 30 days post-operatively [92]. Th ere was no diff erence in clinical scores or healing rates at longer-term follow up. Barber and colleagues [93] showed lower re-tear rates after rotator cuff repair with the use of platelet rich fi brin matrix, but interestingly, there was no diff erence in clinical outcome scores. Currently, use of PRP has marginal clinical support for treatment of rotator cuff repair or cartilage healing [89], and while PRP is a safe treatment for clinical use, its effi cacy remains debatable.

Conclusion
Degenerative conditions of the shoulder remain a signifi cant source of pain and disability in the general and aging populations. Th e burden of arthritis and rotator cuff disease makes them prime topics for basic and translational research. While total shoulder arthroplasty remains the last resort for treating severe glenohumeral disorders, such as OA, other therapies are emerging to aid in improving healing of native tissues. Tendon research has focused on preventing failures of rotator cuff repair and augmenting the biological healing of the rotator cuff . Many potential therapies hold promise, and the implementation of new technologies such as bioengineered scaff olds, novel stem cell sources, and controlled-release growth factors will likely navigate the future of treatment modalities for shoulder pathologies.

Competing interests
The authors do not have any competing interests related to the content of this review.