Reverse total shoulder arthroplasty (RSA) with glenoid bone grafting has become a common option for the management of significant glenoid bone loss and deformity associated with glenohumeral osteoarthritis. Despite the increasing utilization of this technique, our understanding of the rates of bone graft union, complications and outcomes are limited. The objectives of this systematic review are to determine 1) the overall rate of bone graft union, 2) the rate of union stratified by graft type and technique, 3) the reoperation and complication rates, and 4) functional outcomes, including range of motion (ROM) and functional outcome scores following RSA with glenoid bone grafting.

A comprehensive search of MEDLINE, Embase, and CINAHL databases was completed for studies reporting outcomes following RSA with glenoid bone grafting. Inclusion criteria included clinical studies with greater than 10 patients, and minimum follow up of one year. Studies were screened independently by two reviewers and quality assessment was performed using the MINORs criteria. Pooled and frequency-weighted means and standard deviations were calculated where applicable.

Overall, 15 studies were included, including nine retrospective case series (level IV), four retrospective cohort studies (level III), one prospective cohort study (level II) and one randomized control trial (level I). The entire cohort consisted of 555 patients with a mean age of 71.9±2.1 years and 70 percent female. The mean follow-up was 33.8±9.4 months. Across all procedures, 84.9% (N=471) were primary arthroplasties, and 15.1% (N=84) were revisions. The overall graft union rate was 89.2%, but was higher at 96.1% among studies that used autograft bone (9 studies, N=308). When stratified by technique, bone graft for the purposes of lateralization resulted in a 100% union rate (4 studies, N=139), while eccentric bone grafts used in asymmetric bone loss resulted in a lower union rate of 84.9% (10 studies, N=345). The overall revision rate was 6.5%, and was lowest following primary cases at 1.8% (11 studies, N=393). The pooled mean scapular notching rate was 20.1% (12 studies, N=497). Excluding notching, the pooled mean complication rate was 21.5% for all cases and 13% for primary cases (11 studies, N=393). When reported, there was significant improvement in post-operative ROM in all planes. There was also improvement in functional outcome scores, whereby the frequency-weighted mean Constant score increased from 25.9 to 67.2 (8 studies, N=319), ASES score increased from 34.7 to 75.2 (4 studies, N=142), and SST score increased from 2.1 to 7.6 (5 studies, N=196) at final follow up.

This review demonstrates that glenoid bone grafting with RSA results in good mid-term clinical and radiographic outcomes. Union rate appears to depend highly on graft type and technique, whereby the highest union rates were seen following the use of autograft bone for the purposes of lateralization. Interestingly, the union rate of autograft bone for the purposes of augmentation in eccentric bone loss is considerably lower and its impact on the long-term survivorship of the implant remains unknown.

Patient perceptions regarding the functional outcomes and return to sports after shoulder replacement are often pessimistic, with many patients presenting for shoulder replacement surgery after months or years of avoiding the procedure so they could continue to live the current life they have, despite the increasing pain and dysfunction. Less common, but becoming more frequent, patients present with expectations that they will be able to return to all activities including heavy resistance training, cross-fit, rock climbing, and other strenuous overhead sports. In the past, little information has been available regarding the activities of shoulder arthroplasty patients after surgery. Typically, the boundaries have been set by the surgeon, with many patients cautioned or even prohibited from overhead sports, weight training, or heavy work responsibilities. A typical set of guidelines may include no repetitive overhead sports, except for recreational swimming, and no lifting over 20 pounds. Golf, jogging, hiking and other activities are allowed.

The origin of these restricted guidelines and expectations is unknown, but many believe that since the results of shoulder replacements are less favorable in younger patients, it may be due to the overuse or abuse of the shoulder joint that is more typical at a younger age. Others have suggested that common sense prevails and that an artificial joint made of metal and plastic has a finite number of total movements and tolerance to resistance activities, and therefore keeping these activities at a minimum would extend the longevity of the artificial joint. None of these concepts are backed up by evidence-based literature, essentially reflecting the personal bias of the surgeons who care for patients with these problems. Despite all of the sophisticated research, scoring scales, outcome measures, and value-based metrics, the only outcome that really matters is whether the patient can return back to their normal way of life, at home, at work, during sport, or any activity that is important to them.

Recent studies of patients who have had joint replacement surgery have revealed that our patients who participated in sports and work activities before surgery have a strong predilection to returning to those activities after successful shoulder replacement. The most common sports that shoulder arthroplasty patients enjoy including golf, swimming, tennis, but may also include many other choices including fitness activities, rowing, skiing, basketball, and softball. As expected, the return to these sports is less for reverse shoulder arthroplasty patients vs. anatomic shoulder arthroplasty patients. In a systematic review, more than 90% of anatomic shoulder replacement patients returned to sport, while 75% of reverse shoulder arthroplasty patients returned to some sporting activity. This may reflect the constraints of the reverse prosthesis, or, quite possibly reflect the typically older age and more sedentary lifestyle of patients who are indicated for reverse shoulder arthroplasty. In addition, if the patient had a preoperative expectation of return to recreational and sports activities as part of their normal way of life, their final results demonstrated better overall outcomes.

Shoulder arthroplasty surgeons should be concerned about the outcomes desired by our patients, and the results that provide true value to their lives. We are now more aware of the activities that they are going to return to, whether we recommend restrictions or not.

The advent of modern anatomic shoulder arthroplasty occurred in the 1990's with the revelation that the humeral head dimensions had a fixed ratio between the head diameter and height. As surgeons moved from the concept of balancing soft tissue tension by using variable neck lengths for a given humeral head diameter, a flawed concept based on lower extremity reconstruction, improvements in range of motion and function were immediately observed. Long term outcome has validated this guiding principle for anatomic shoulder replacement with improved longevity of implants, improved patient and surgeon expectations and satisfaction with results.

Once the ideal humeral head prosthesis is identified, and its position prepared, the surgeon must use a method to fix the position of the head that is correct in three dimensions and has the security to withstand patient activities and provide maximal longevity. Based again on lower extremity concepts, long stems were the standard of care, initially with cement, and now, almost universally without cement for a primary shoulder replacement. The incredibly low revision rates for humeral stem aseptic loosening shifted much of the attempted innovation to the challenges on the glenoid side of the reconstruction.

However, glenoid problems including revision surgery, infections, periprosthetic fractures, and other complications often required the removal of the humeral stem. And, in many cases, the overall results of the procedure and the patient's long-term outcome was affected by the difficulty in removing the stem, leading surgeons to compromise the revision procedure, avoid revision surgery, or add to the overall morbidity with humeral fractures and substantial bone loss.

With improved technology, including bone ingrowth methods, better matching of the proximal stem geometry to the humerus, and an understanding that the center of rotation (torque) on the humeral component is at the level of the humeral osteotomy, shorter stems and stemless humeral components were developed, now more than 10 years ago, primarily in Europe.

With more than a decade of experience, our European colleagues have shown us that stemless humeral component replacement with a device that has both cortical and cancellous fixation is as effective as a stemmed device, easier to implant as well as revise when needed. The short-term results of the cancellous fixation stemless devices are acceptable, but longer follow-up is needed.

Currently, the most widely used humeral components in the USA are short stem components, although the recent FDA approval of numerous stemless devices has initiated a shift from short stems to stemless devices. The truth is, short stem devices have a firm position in the USA surgeons' armamentarium today due to regulatory restrictions. A decade ago, without a predicate on the market, it was not conceivable that a stemless device that was already gaining popularity in Europe would be able to get 510K approval, and therefore would require a lengthy and expensive FDA IDE process. However, shorter stems had already been approved in the USA, as long as the stem length was 7 centimeters, matching the market predicate.

Now, in 2018, based on evidence and outcomes, stemless humeral components should be the first choice when treating primary osteoarthritis of the glenohumeral joint. Short stem or longer stem devices should be reserved for those cases where stemless fixation is not possible, which is less than 10% of patients with primary OA of the shoulder.

Shoulder arthritis in the young adult is a deceptive title. The literature is filled with articles that separate outcomes based on an arbitrary age threshold and attempt to provide recommendations for management and even potential criteria for implanting one strategy over another using age as the primary determinant. However, under the age of 50, as few as one out of five patients will have arthritis that can be accurately classified as osteoarthritis. Other conditions such as post-traumatic arthritis, post-surgical arthritis including capsulorrhaphy arthropathy, and rheumatoid arthritis create a mosaic of pathologic bone and soft tissue changes in our younger patients that distort the conclusions regarding “shoulder arthritis” in the young adult. In addition, we are now seeing more patients with unique conditions that are still poorly understood, including arthritis of the pharmacologically performance-enhanced shoulder.

Early arthritis in the young adult is often recognised at the time of arthroscopic surgery performed for other preoperative indications. Palliative treatment is the first option, which equals “debridement.” If the procedure fails to resolve the symptoms, and the symptoms can be localised to an intra-articular source, then additional treatment options may include a variety of cartilage restoration procedures that have been developed primarily for the knee and then subsequently used in the shoulder, including microfracture, and osteochondral grafting. The results of these treatments have been rarely reported with only case series and expert opinion to support their use.

When arthritis is moderate or severe in young adults, non-arthroplasty interventions have included arthroscopic capsular release, debridement, acromioplasty, distal clavicle resection, microfracture, osteophyte debridement, axillary nerve neurolysis, and bicep tenotomy or tenodesis, or some combination of these techniques. Again, the literature is very limited, with most case series less than 5 years of follow-up. The results are typically acceptable for pain relief, some functional improvement, but not restoration to completely normal function from the patient's perspective.

Attempts to resurface the arthritic joint have resulted in limited benefits over a short period of time in most studies. While a few remarkable procedures have provided reasonable outcomes, they are typically in the hands of the developer of the procedure and subsequently, other surgeons fail to achieve the same results. This has been the case with fascia lata grafting of the glenoid, dermal allografts, meniscal allografts, and even biologic resurfacing with large osteochondral grafts for osteoarthritis. Most surgical interventions that show high value in terms of improvement in quality of life require 10-year follow-up. It is unlikely that any of these arthroscopic procedures or resurfacing procedures will provide outcomes that would be valuable in terms of population healthcare; they are currently used on an individual basis to try to delay progression to arthroplasty, with surgeon bias based on personal experience, training, or expert opinion.

Arthroplasty in the young adult remains controversial. Without question, study after study supports total shoulder arthroplasty over hemiarthroplasty once the decision has been made that joint replacement is the only remaining option.

Patients over 70 years old have subclinical or impending rotator cuff dysfunction, raising concern about TSA in this population. The purpose of this study is to examine whether reverse total shoulder arthroplasty (RTSA) should be considered for the treatment of glenohumeral osteoarthritis in the presence of an intact rotator cuff (GHOA+IRC in patients older than 70 years of age.

Twenty-five elderly (>70 years) patients at least one year status-post RTSA for GHOA+IRC were matched via age, sex, body mass index, smoking status, and whether the procedure involved the dominant extremity with 25 GHOA+IRC patients who received anatomic total shoulder arthroplasty (TSA). Standardised outcome measures, range of motion, and treatment costs were compared between the two groups. Treatment cost was assessed using implant and physical therapy costs as well as reimbursement.

Patients who received RTSA for GHO+IRC had significantly lower pre-operative active forward elevation (AFE, 69° vs. 98°, p <0.001) and experienced a greater change in AFE (p=0.01), but had equivalent AFE at final follow-up (140° vs. 142°, p=0.71). Outcomes were otherwise equivalent between groups with no differences. In both those patients who underwent TSA and those that underwent RTSA, significant improvements between pre-operative and final follow-up were seen in all standardised outcome measures and in AFE (p<0.001 in all cases). RTSA provided these outcomes at a cost savings of $2,025 in Medicare reimbursement due to decreased physical therapy costs.

In patients over the age of 70 with GHOA+IRC, RTSA provides similar improvement in clinical outcomes to TSA at a reduced cost while avoiding issues related to the potential for subclinical or impending rotator cuff dysfunction.

Glenohumeral osteoarthritis (OA) is a challenging clinical problem in young patients. Given the possibility of early glenoid component loosening in this population with total shoulder arthroplasty (TSA), and subsequent need for early revision, alternative treatment options are often recommended to provide pain relief and improved range of motion. While nonoperative modalities including nonsteroidal anti-inflammatory medications and physical therapy focusing on rotator cuff strengthening and scapular stabilization may provide some symptomatic relief, young patients with glenohumeral OA often need surgery for improved outcomes. Joint preserving techniques, such as arthroscopic debridement with removal of loose bodies and capsular release, with or without biceps tenotomy or tenodesis, remains a viable nonarthroplasty option in these patients.

Clinical studies evaluating the outcomes of arthroscopic debridement for glenohumeral OA in young patients have had favorable outcomes. Evidence suggests that earlier stages of glenohumeral OA have more favorable outcomes with arthroscopic debridement procedures, with worse outcomes being observed in patients with complete joint space loss and bipolar chondral lesions. More advanced arthroscopic options include inferior osteophyte excision and axillary neurolysis or microfracture of chondral lesions, both of which have demonstrated favorable early clinical outcomes.

Patients with some preserved joint space and small osteophytes can avoid arthroplasty and have improved functional outcomes after arthroscopic debridement for glenohumeral OA. Caution should be advised when indicating this procedure for patients with large osteophytes, grade IV bipolar lesions, biconcave glenoids, and complete loss of joint space.

Humeral resurfacing arthroplasty has been advocated as an alternative to stemmed humeral component designs given its ability to preserve proximal bone stock. Further, these implants have become more attractive given the possibility of stem-related complications including humeral fracture, stress shielding, and osteolysis; complications that may necessitate fixation, revision to long stem components, or reverse total shoulder arthroplasty. As more total shoulder arthroplasties are performed in younger patient populations, the likelihood of increased revision procedures is inevitable. Maintaining proximal bone stock in these cases with use of a resurfacing arthroplasty not only facilitates explant during revision arthroplasty, but preservation of proximal metaphyseal bone facilitates reimplantation of components. Clinical results of these resurfacing components have demonstrated favorable results similar to stemmed designs.

Unfortunately, resurfacing arthroplasty may not be as ideal as was hoped with regard to recreating native humeral anatomy. Further, resurfacing arthroplasty may increase the risk of peri-prosthetic humeral fracture, and lack of a formal humeral head cut makes glenoid exposure more difficult, which may be associated with a higher degree of neurovascular injury. Stemless humeral components are designed for strong metaphyseal fixation and avoid the difficulty with glenoid exposure seen in resurfacing designs, as these components require a formal humeral head cut.

Early clinical outcomes of a single stemless design demonstrated significant improvements in clinical outcome scores, without evidence of component migration, subsidence or loosening. The only mid-term clinical results of stemless design implants are seen with the Arthrex Eclipse system (Arthrex, Naples, FL). In a prospective study involving 78 patients at 5-year follow-up, significant improvements were observed in clinical outcome scores. While there was evidence of proximal stress shielding in an older population, this did not influence shoulder function. The overall revision rate was 9% at 5 years, with no component necessitating revision as a result of humeral component loosening.

Resurfacing arthroplasty and stemless humeral components in total shoulder arthroplasty remain attractive options to preserve proximal metaphyseal bone stock, avoiding stem-related complications. Early and mid-term clinical outcomes are comparable to stemmed designs and demonstrate no evidence of humeral component loosening.

The incorporation of platelet rich plasma (PRP) in the treatment of various musculoskeletal conditions has increased exponentially over the past decade. While described most often as an augment or treatment for tendinopathies and acute tendon injuries, more recently, PRP has been described as an adjunct to arthroplasty procedures, mostly with respect to knee arthroplasty. In the shoulder, only a single study has been published, in which Zavadil and colleagues performed a randomised study of 40 patients undergoing total shoulder arthroplasty undergoing either treatment with autologous platelet gel and platelet poor plasma (n=20) or undergoing no biologic treatment (control group, n=20). The authors noted that the treatment group had significantly lower pain scores, less pain medication requirements, and improved internal rotation when compared to controls; in addition, there were no significant differences in post-operative (compared to pre-operative) hemoglobin levels or length of stay. The vast majority of arthroplasty studies discussing PRP analyze the impact of treatment on wound healing, post-operative pain, post-operative range of motion, and need for post-operative blood transfusions. Unfortunately, due to the substantial variability of methodology (not all PRP preparations are the same) in the available studies as well as the variability in outcomes reporting, direct comparison between different studies is not feasible. Here, we discuss the basic science elements of PRP relevant to arthroplasty, the variability of PRP solutions, the specific applications of PRP in arthroplasty, and the latest clinical outcomes analyses of patients undergoing PRP therapy in conjunction with shoulder arthroplasty.

Total shoulder arthroplasty (TSA) is an excellent surgical solution for patients with shoulder arthritis, providing good to excellent results in the vast majority of patients. Complications are rare, however, when they occur, can be devastating for both the patient and surgeon. An increasingly recognised complication of TSA is glenoid component failure. In a recent review of nearly 4000 shoulder arthroplasties, symptomatic glenoid component loosening occurs at a rate of 1.2% per year, while asymptomatic radiolucent lines occur at a rate of 7.3% per year. In addition, keeled glenoid components have been found to have an increased incidence of radiolucent lines compared to pegged glenoid components at both short and longer-term follow-up. Further, pegged glenoid components are associated with a lower risk of revision arthroplasty compared to keeled glenoid components. In a separate study of approximately 4600 shoulder arthroplasties, metal-backed glenoid components were found to have significantly lower rates of radiolucent lines, radiographic loosening, and component failure compared to all-polyethylene glenoid components. Despite these findings, however, metal-backed components are significantly more likely to require revision surgery (3:1) compared to all-polyethylene components. For the failed TSA due to glenoid component failure, revision with glenoid reimplantation may be considered, but is associated with a high rate of recurrent glenoid loosening. Alternatively, revision to a reverse shoulder arthroplasty, is feasible, but is technically demanding with high complication and reoperation rates. This paper will discuss the etiology, work-up, and treatment of patients with glenoid component failure following TSA.

Total shoulder arthroplasty (TSA) and reverse shoulder arthroplasty (RSA) are excellent surgical options for individuals with shoulder arthritis, providing good to excellent results in the vast majority of patients. Complications are rare, but can be devastating for both the patient and surgeon. An uncommon, but extremely problematic complication following shoulder arthroplasty is shoulder stiffness. While substantial literature discussing post-arthroplasty stiffness is available for other joints such as the hip, knee, and elbow, there is a paucity of research available discussing this complication in the shoulder. As noted in multiple reviews, diminished range of motion following TSA or RSA may be due to a number of factors, including pre-operative diagnosis of proximal humerus fracture, inadequate post-operative rehabilitation, implant-related factors such as malpositioning and/or inappropriate-sized implants, and heterotopic ossification. Often, pathology leading to post-arthroplasty stiffness involves scarring of the long head of the biceps tendon, rotator cuff impingement, as well as cuff tendonitis. Periprosthetic joint infection (PJI) is also important to recognise, and may be difficult to diagnose, especially in cases of Propionibacterium acnes infections. Importantly, PJI may present with stiffness as well as instability, and thus a high index of suspicion with a low threshold to aspirate is necessary in these challenging patients. Treatment of patients with stiffness following arthroplasty is challenging, and may involve arthroscopic intervention with or without manipulation, as well as manipulation under anesthesia alone. This paper will discuss the etiology, work-up, and treatment of patients with shoulder stiffness following TSA and RSA.

The Hill-Sachs lesion is a bony defect of the humeral head that occurs in association with anterior instability of the glenohumeral joint. Hill-Sachs lesions are common, with an incidence approaching nearly 100% in the setting of recurrent anterior glenohumeral instability. However, the indications for surgical management are very limited, with less than 10% of anterior instability patients considered for treatment of the Hill-Sachs lesion. Of utmost importance is addressing bone loss on the anterior-inferior glenoid, which is highly successful at preventing recurrence of instability even with humeral bone loss. In the rare situation where the Hill-Sachs lesion may continue to engage the glenoid, surgical management is indicated. Surgical strategies are variable, including debridement, arthroscopic remplissage, allograft transplantation, surface replacement, and arthroplasty. Given that the population with these defects is typically comprised of young and athletic patients, biologic solutions are most likely to be associated with decades of sustainable joint preservation, function, and stability. The first priority is maximizing the treatment of anterior instability on the glenoid side. Then, small lesions of less than 10% are ignored without consequence. Lesions involving 10–20% of the humeral head are treated with arthroscopic remplissage (defect filled with repair of capsule and infraspinatus). Lesions greater than 20% that extend beyond the glenoid tract are managed with fresh osteochondral allografts to biologically restore the humeral head. Lesions great than 40% are most commonly associated with advanced arthritis and deformity of the humeral articular surface and are therefore treated with a humeral head replacement. This treatment algorithm maximises our ability to stabilise and preserve the glenohumeral joint.

The functional importance of the long head of biceps tendon remains controversial, but it is well accepted as an important source of anterior shoulder pain. Both biceps tenotomy and tenodesis have well-documented results in the native shoulder. Management of the long head of biceps tendon during shoulder arthroplasty remains controversial. The existing literature supports surgical treatment of the biceps during shoulder arthroplasty. Walch et al. reported the largest series in multicenter study of over eight hundred shoulder arthroplasty patients with or without biceps tenodesis. The authors found more reproducible pain relief with biceps tenodesis and no difference in range of motion. Similarly, Soliman et al. reported on a prospective review of 37 patients undergoing hemiarthroplasty for fracture randomly assigned to biceps tenodesis vs. no treatment. The authors found a statistically significant improvement in Constant score and shoulder pain with biceps tenodesis. If left untreated during shoulder arthroplasty, the intact biceps tendon may be a source of anterior shoulder pain requiring revision surgery. Tuckman et al. reported excellent pain relief after biceps tenotomy or tenodesis for biceps-related pain after previous shoulder arthroplasty. The decision to perform a biceps tenotomy versus a tenodesis during shoulder arthroplasty also remains controversial. Tenotomy may increase the risk of cosmetic pop-eye deformity and muscle cramping or fatigue over tenodesis. Therefore, routine long head of biceps soft tissue tenodesis is recommended during shoulder arthroplasty as it safe, reproducible, cost-effective, associated with improved outcome scores, and minimises the risk of cosmetic deformity and pain associated with biceps tenotomy.

The Hill-Sachs lesion is a bony defect of the humeral head that occurs in association with anterior instability of the glenohumeral joint. Hill-Sachs lesions are common, with an incidence approaching nearly 100% in the setting of recurrent anterior glenohumeral instability. However, the indications for surgical management are very limited, with less than 10% of anterior instability patients considered for treatment of the Hill-Sachs lesion. Of utmost importance is addressing bone loss on the anterior-inferior glenoid, which is highly successful at preventing recurrence of instability even with humeral bone loss. In the rare situation where the Hill-Sachs lesion may continue to engage the glenoid, surgical management is indicated. Surgical strategies are variable, including debridement, arthroscopic remplissage, allograft transplantation, surface replacement, and arthroplasty. Given that the population with these defects is typically comprised of young and athletic patients, biologic solutions are most likely to be associated with decades of sustainable joint preservation, function, and stability. The first priority is maximising the treatment of anterior instability on the glenoid side. Then, small lesions of less than 10% are ignored without consequence. Lesions involving 10–20% of the humeral head are treated with arthroscopic remplissage (defect filled with repair of capsule and infraspinatus). Lesions greater than 20% that extend beyond the glenoid tract are managed with fresh osteochondral allografts to biologically restore the humeral head. Lesions great than 40% are most commonly associated with advanced arthritis and deformity of the humeral articular surface and are therefore treated with a humeral head replacement. This treatment algorithm maximises our ability to stabilise and preserve the glenohumeral joint.