This annual update on shoulder and elbow surgery is a review of the most relevant studies from July 2007 through June 2008. It includes clinical and basic-science articles from The Journal of Bone and Joint Surgery (American Volume), The Journal of Bone and Joint Surgery (British Volume), the Journal of Shoulder and Elbow Surgery, and Arthroscopy: The Journal of Arthroscopic and Related Surgery. Relevant level-I and II studies from other medical journals are included where appropriate.
The level of evidence is indicated at the end of each review when it is known. Particular attention should be paid to the level-I and II studies because they represent randomized controlled studies. Additional level-III and IV studies representing important topics in shoulder and elbow surgery are also included in the review.
It should be noted, however, that not all level-I or II studies are well designed, well executed, or scientifically valid. Interestingly, Cowan et al.1 reviewed level-I and II therapeutic studies on lateral epicondylitis to measure variation in quality. Fifty-four prospective, randomized therapeutic trials were evaluated according to the Oxford Centre for Evidence-based Medicine Levels of Evidence, a modification of the Coleman Methodology Score, and the revised CONSORT (Consolidated Standards of Reporting Trials) score. Deficiencies were noted in many studies with respect to study design, patient enrollment, study execution, and data analysis. These results point out that level-I and II studies require the same critical analysis as do studies with a lower level of evidence.
Frozen Shoulder
The nonoperative treatment of adhesive capsulitis often includes intra-articular corticosteroid injection. However, there is debate over the use of single or multiple injections. Shah and Lewis2 performed a systematic review of randomized controlled trials. Nine randomized controlled trials were identified, and four studies were rated as high quality. Three high-quality studies showed benefit in association with the use of multiple corticosteroid injections as measured on the basis of pain reduction, improved function, and increased range of motion. The evidence suggested that multiple injections were beneficial until sixteen weeks after the date of the first injection. As many as three injections were beneficial, with limited evidence that four to six injections were beneficial. No evidence was found to support giving more than six injections. (Level I)
In addition to intra-articular injections, a home exercise program is often employed. After the failure of nonoperative treatment, manipulation with the patient under anesthesia is often considered. In the randomized blinded trial by Kivimäki et al.3, manipulation with home exercises was compared with an isolated home exercise program for the treatment of frozen shoulder. Shoulder pain and the ability to perform work activities were similar at all time points. The manipulation group demonstrated only slightly better range of motion in comparison with the control group. The authors concluded that manipulation did not offer any advantage over exercises alone. (Level I)
Impingement Syndrome/Rotator Cuff Tendinitis
Subacromial injection of corticosteroids is commonly used for the treatment of impingement syndrome. However, the most accurate method of injection is not well established. Kang et al.4 performed a randomized prospective study to assess the accuracy of subacromial injections. Patients randomly received a subacromial injection of corticosteroids, local anesthetic, and contrast medium from an anterolateral, lateral, or posterior injection site. The accuracy of the injection was confirmed radiographically, and clinical outcomes were measured. The overall accuracy of injection was 70%, with no difference being noted among the three injection sites or in association with body mass index. Accuracy of the injection reliably produced a positive impingement test. Overall, the clinical outcome significantly improved. However, the clinical improvement did not correlate with the accuracy of the injection. (Level I)
In addition to subacromial injections, a resistance training program is often utilized. Lombardi et al.5 performed a randomized study in which patients were assigned to a progressive resistance program or a control group for the treatment of subacromial impingement. Before and after treatment, patients were evaluated with regard to pain, function, quality of life, muscle strength, and the number of anti-inflammatory drugs and analgesics taken. There was significantly greater improvement in terms of pain (both at rest and with activity) and function for patients who performed a resistance program as compared with those in the control group. This study demonstrates that a progressive resistance training program is an important treatment modality for patients with shoulder impingement syndrome. (Level I)
When nonoperative treatment fails in a patient with impingement syndrome, subacromial decompression is often required. At the time of surgery, regional anesthesia provides the added benefit of postoperative pain relief. However, many patients find long-duration complete anesthesia of the arm to be uncomfortable. Nisar et al.6 performed a randomized study in which interscalene block was compared with subacromial bursa block, with general anesthesia being used as the control. The postoperative pain scores in the interscalene block and subacromial bursa block groups were better than those in the control group in the first twelve hours postoperatively. The patients in the control group consumed significantly more morphine compared with those in the subacromial bursa block group and the interscalene block group. The oral analgesic intake was significantly less in the subacromial bursa block and interscalene block groups than in the control group. The authors concluded that a subacromial bursa block provides effective, safe, and easily administered postoperative analgesia for patients undergoing arthroscopic subacromial decompression. (Level I)
The surgical treatment of impingement syndrome traditionally has been subacromial decompression. Taverna et al.7 performed a randomized prospective study to determine whether radiofrequency microtenotomy was effective for treating chronic supraspinatus tendinosis when compared with arthroscopic subacromial decompression. All patients had had a failure of nonoperative treatment and had Neer stage-II impingement syndrome, radiographic evidence of a type-II acromion, and magnetic resonance imaging or ultrasound evidence of supraspinatus tendinosis. A significant reduction in pain and improvement in function were observed in both groups following surgery. (Level I)
Acromioclavicular Joint
The most appropriate treatment for acute Rockwood type-III acromioclavicular joint separations is controversial. In the retrospective study by Gstettner et al.8, patients with acute type-III acromioclavicular separations that had been treated surgically with hook plate fixation were compared with those who had been managed conservatively. At the time of the latest follow-up, the surgically treated group had a significantly better average Constant score in comparison with the conservatively treated group. The mean coracoclavicular distance was significantly greater in the conservatively treated group as compared with the surgically treated group (15.9 compared with 12.1 mm). In this retrospective study, better clinical and radiographic results were achieved by means of surgical treatment with the hook plate as compared with conservative treatment. (Level IV)
Rotator Cuff
The natural history of massive rotator cuff tears has not been clearly defined. Zingg et al.9 followed a group of patients who had been managed nonoperatively with two and three-tendon rotator cuff tears. All patients were evaluated on the basis of a clinical examination, standard radiographs, and magnetic resonance imaging. The active range of motion did not change over time. Notable findings included progression of glenohumeral osteoarthritis, a decrease in the acromiohumeral distance, an increase in the size of the tear, and an increase in fatty infiltration by approximately one stage in all three rotator cuff muscles. Four of the eight rotator cuff tears that were graded as reparable at the time of the diagnosis became irreparable. The authors concluded that patients with a nonoperatively treated, massive rotator cuff tear can maintain satisfactory shoulder function in the short term despite substantial progression of degenerative structural joint changes. With tear progression, there is a risk that a reparable tear will become irreparable. (Level IV)
Open rotator cuff repair remains the gold standard for comparison of different surgical techniques. Mohtadi et al.10 performed a randomized trial in which open rotator cuff repair was compared with arthroscopic acromioplasty combined with mini-open rotator cuff repair. The outcomes within the first three months postoperatively and at three months of follow-up were significantly better for the mini-open repair group as compared with the open rotator cuff repair group. However, these early differences in outcome were not significant between the treatment groups at the time of the latest follow-up (average, twenty-eight months). Patient outcomes improved from baseline at all postoperative measurement intervals. (Level I)
Arthroscopic rotator cuff repair techniques continue to evolve, and the best method has not been clearly established. Franceschi et al.11 performed a randomized controlled trial in which arthroscopic single-row suture anchor repair was compared with arthroscopic double-row suture anchor repair for the treatment of full-thickness rotator cuff tears. At the time of the two-year follow-up, no significant differences were seen with respect to clinical outcome or range of motion. Postoperative magnetic resonance arthrography showed no difference between the treatment groups with regard to the integrity of the repair. This study indicates that single and double-row techniques provide comparable clinical outcome at two years. (Level I)
The role that rotator cuff healing plays with regard to the time needed to achieve significant functional recovery following surgery is not well understood. Charousset et al.12 conducted a prospective study to evaluate the relationship between rotator cuff healing and the time required to achieve significant postoperative functional recovery. The Constant score significantly improved from the third month after surgery until the twelfth month, after which it stabilized. Female sex, heavy work involving the upper extremity, poor bone quality, and absence of healing were negative predictive factors. In particular, the absence of tendon healing led to a worsening of the results from the twelfth month onward. (Level IV)
Long-standing rotator cuff tears are associated with atrophy and fatty infiltration of the musculature. The integrity of the rotator cuff repair following surgery may have an effect on these findings. Gerber et al.13 prospectively correlated postoperative muscular atrophy and fatty infiltration with the integrity of the rotator cuff repair and strength. Postoperative magnetic resonance imaging established the integrity of the repair and provided a comparison with preoperative magnetic resonance imaging findings to determine the extent of atrophy and fatty infiltration. Within one year after successful tendon repair, fatty infiltration did not recover, and atrophy improved partially at best. If the rotator cuff repair failed, atrophy and fatty infiltration progressed significantly.
Suprascapular neuropathy has been associated with rotator cuff tears with substantial retraction. Costouros et al.14 performed a retrospective study to determine if arthroscopic rotator cuff repair, even if partial, was associated with reversal of the neuropathy and clinical improvement. All patients in the study had a massive rotator cuff tear and an electrodiagnostic evaluation documenting an isolated suprascapular neuropathy. In the patients who underwent either partial or complete arthroscopic repair, follow-up electrodiagnostic evaluation, performed six months postoperatively, demonstrated partial or full recovery of the nerve palsy. Nerve recovery correlated with complete pain relief and marked improvement in function. Suprascapular nerve dysfunction should be considered in patients with a large, retracted rotator cuff tear. If documented preoperatively, the presence of nerve dysfunction should be a consideration in surgical decision-making. (Level IV)
Irreparable posterosuperior rotator cuff tears can result in loss of active external rotation, severely impacting an individual's ability to perform activities of daily living. In order to restore active external rotation, Boileau et al.15 developed a modified L'Episcopo procedure in which the latissimus dorsi and teres major were transferred and reattached laterally through a single deltopectoral approach. The transfer was performed in patients with an isolated loss of active external rotation related to an irreparable posterosuperior rotator cuff tear. The average increase in active elevation was 34.7°, and the average increase in active external rotation was about 27°. All patients but one were satisfied or very satisfied with the result, with the major contributor to their satisfaction being the ability to control the spatial positioning of the arm. When the modified transfer is performed with a reverse shoulder arthroplasty, it allows restoration of both active elevation and external rotation.
Superior Labral (SLAP) Tears
The diagnosis of SLAP (superior labrum anterior posterior) tears remains difficult. Pandya et al.16 studied the sensitivity of physical examination, magnetic resonance imaging, and/or magnetic resonance arthrography for the identification of arthroscopically confirmed SLAP lesions of the shoulder. Before undergoing surgery, all patients underwent a standardized physical examination and had a magnetic resonance imaging scan and/or magnetic resonance arthrogram. The sensitivities of clinical tests were 90% for the O'Brien test, 80% for the Mayo shear test, and 76% for the Jobe relocation test. The sensitivity of a physical examination with any one of these three SLAP provocative tests being positive was 100%. Magnetic resonance imaging and magnetic resonance arthrogram imaging had lower sensitivity than these physical examination tests for the diagnosis of SLAP lesions. The authors concluded that patient history, demographic characteristics, and the physical examination by the surgeon should remain central to the diagnosis of SLAP lesions. (Level II)
Superior labral tears have a known association with paralabral cysts encroaching on the spinoglenoid notch. The treatment of symptomatic spinoglenoid cysts has included observation, needle aspiration, open excision, and arthroscopic decompression. Schroder et al.17 performed a study to assess whether labral repair alone would lead to cyst resolution and pain relief. Postoperative magnetic resonance imaging demonstrated complete resolution of the cyst in 88% of the forty-two patients. In five patients, a cyst was still present but was reduced in size. These five patients had resolution of pain and were satisfied with shoulder function. Three patients with preoperative muscular atrophy without fatty infiltration regained normal-appearing muscle, whereas the seven patients who had fatty changes preoperatively continued to demonstrate those changes postoperatively. A limitation of this study was the absence of preoperative or postoperative electrodiagnostic testing. This limitation prevents comment regarding the resolution of suprascapular neuropathy with an isolated labral repair. However, the authors found that most spinoglenoid cysts resolve and that patient satisfaction can be expected to be high after labral fixation without cyst decompression. (Level IV)
Glenohumeral Instability
Hovelius et al.18 performed a prospective multicenter study of 229 shoulders to evaluate the results of nonoperative treatment of primary anterior shoulder dislocation after twenty-five years. Half (116) of the shoulders had no recurrent dislocations or had only one recurrent dislocation. Thirty-three recurrent dislocations (representing 14.4% of the 229 shoulders) became stable over time, and eighteen (representing 7.9% of the shoulders) were considered to be still recurrent. Associated fractures of the greater tuberosity were present in 10% of the shoulders. Immobilization after the primary dislocation did not appear to change the prognosis. With the numbers available for study, the sex and level of athletic activity of the patients did not appear to affect the redislocation rate. However, women had significantly worse DASH (Disabilities of the Arm, Shoulder and Hand) scores than did men. The authors concluded that half of the primary anterior shoulder dislocations that had been treated nonoperatively in patients twelve to twenty-five years of age had not recurred or had become stable over time. (Level I)
The decision to operate on a patient with traumatic anterior shoulder instability is based, in part, on the likelihood of future dislocations. However, as noted in the study above, not all patients experience recurrent instability and some who do learn to cope with it. Sachs et al.19 performed a natural history study to determine whether it is possible to predict the need for future surgery at the time of the initial injury. Of the 131 patients who were enrolled in the study, forty-three (33%) had recurrent instability and twenty-nine (22%) requested surgery. The majority of redislocations occurred in patients under the age of forty years who either participated in contact or collision sports or used the arm at or above chest level in their occupation. Interestingly, only half of the patients with recurrent instability underwent surgery. The outcome scores for patients without recurrent instability were high and were equivalent to those for the cohort of patients who had had a successful Bankart repair. Patients who had redislocation but chose to cope with the instability rather than to have surgery had lower outcome scores. According to the results of this study, the decision to proceed with surgery on the basis of the presumption of future dislocations, unhappiness, and disability cannot be justified. (Level I)
Previous studies by Itoi have demonstrated that the Bankart lesion is reduced against the glenoid rim better with the arm in external rotation than it is with the arm in internal rotation. In their recent randomized controlled trial, Itoi et al.20 sought to determine the benefit of three weeks of immobilization in external rotation. Recurrent instability was the primary outcome measure. An intention-to-treat analysis revealed that the recurrence rate in the external rotation immobilization group was significantly lower than that in the internal rotation immobilization group (26% compared with 42%). A relative risk reduction of 38.2% was demonstrated for all age groups in association with immobilization in external rotation. In the subgroup of patients who were thirty years of age or younger, the relative risk reduction was 46.1%. The authors concluded that immobilization in external rotation appears to be particularly beneficial for patients who are thirty years of age or younger. (Level II)
Large, displaced fractures of the glenoid associated with traumatic anterior dislocations are frequently treated with surgical fixation. Maquieira et al.21 evaluated a group of patients who had nonoperative treatment of anteroinferior glenoid rim fractures measuring >5 mm in size with >2 mm of displacement. Nonoperative treatment was considered only if post-reduction radiographs showed the humeral head to be centered on the glenoid. After an average duration of follow-up of 5.6 years, the mean Constant score and subjective shoulder value were 98% and 97%, respectively. There were no redislocations or subluxations, and the apprehension test was consistently negative. All fragments healed with an average intra-articular step of 3.0 mm. The authors concluded that nonoperative treatment of large, displaced glenoid rim fractures can be considered only if the glenohumeral joint remains concentrically reduced.
When surgery is required for the treatment of shoulder instability, both arthroscopic and open techniques are available. A perceived advantage of arthroscopic as compared with open stabilization is that the subscapularis muscle-tendon unit is not violated. Hiemstra et al.22 performed a randomized controlled trial to evaluate shoulder strength following arthroscopic stabilization and open shoulder stabilization through a subscapularis splitting approach. Patients underwent isokinetic strength testing, which was normalized to body mass. No significant differences were identified between operative groups with regard to internal rotation strength at one year after surgery. The authors did not evaluate open stabilization surgery performed through the more commonly performed subscapularis tenotomy. (Level I)
When considering an arthroscopic Bankart repair, the ability to preoperatively identify patients who are at risk for recurrent instability would be tremendously valuable for selecting an alternative surgical procedure. Balg and Boileau23 developed a 10-point preoperative instability severity index score utilizing a set of risk factors. The risk factors included a patient age of less than twenty years at the time of surgery, involvement in competitive or contact sports, shoulder hyperlaxity, a Hill-Sachs lesion, and/or loss of the sclerotic inferior glenoid contour. The authors performed a prospective case-control study of consecutive patients who underwent arthroscopic stabilization with use of suture anchors. The instability severity index score was retrospectively applied to the patient population, and it was determined that a score of >6 points carried a significantly increased risk of recurrence. The authors concluded that an arthroscopic Bankart repair is contraindicated for patients with an instability severity index of >6 points, and they favored an open repair (Bristow-Latarjet procedure) for these patients.
Arthroscopic techniques for the treatment of shoulder instability are evolving rapidly. The ultimate measure of success requires comparison with open stabilization, which remains the gold standard. However, meaningful comparisons are difficult because arthroscopic techniques change so quickly. Hobby et al.24 performed a systematic review of studies published between January 1985 and February 2006 to identify those that evaluated the results of arthroscopic procedures for the treatment of chronic anterior shoulder instability or that compared arthroscopic and open surgery. A meta-analysis of appropriate studies was performed in order to summarize risk estimates. The failure rate of arthroscopic shoulder stabilization with use of staples or transglenoid suture techniques appeared to be significantly higher than that of either open surgery or arthroscopic stabilization with use of suture anchors or bioabsorbable tacks. The authors concluded that the rate of failure of arthroscopic anterior stabilization with use of suture anchors or bioabsorbable tacks was similar to that of open stabilization after two years. (Level I)
Robinson et al.25 performed a double-blind, clinical trial of young patients who had sustained a primary anterior glenohumeral dislocation. The goal of the study was to assess the efficacy of a primary arthroscopic Bankart repair. Patients were randomized to treatment with either an arthroscopic examination and joint lavage alone (control) or a Bankart repair. The risk of a further dislocation was reduced by 76% and the risk of all recurrent instability was reduced by 82% in the Bankart repair group as compared with the control group. The functional scores and patient satisfaction were significantly better after arthroscopic repair. Prevention of further instability appeared to account for the improved function associated with Bankart repair as stable shoulders had similar function, irrespective of the initial treatment allocation. The authors concluded that primary arthroscopic repair of a Bankart lesion provides a marked treatment benefit as compared with lavage alone, largely as a result of improved stability of the shoulder. (Level I)
Anterior shoulder instability combined with bone loss may require a surgical procedure to reconstitute glenoid bone stock. An accurate method of measuring glenoid bone loss is essential for determining which surgical procedure best fits the clinical situation. Chuang et al.26 evaluated the ability of three-dimensional computed tomography scans of the glenoid to accurately quantify glenoid bone loss. They defined the glenoid index as the ratio of the maximum inferior diameter of the injured glenoid to that of the uninjured, contralateral glenoid as calculated from the three-dimensional computed tomography scans. If the glenoid index was >0.75, the patient was predicted to benefit from an arthroscopic Bankart repair. However, if the glenoid index was =0.75, the patient was predicted to benefit from an open Latarjet procedure. The three-dimensional computed tomography scans accurately supported the decisions to perform an arthroscopic Bankart repair or an open Latarjet procedure. Surgeons therefore can use the three-dimensional computed tomography scan as an additional diagnostic tool for preoperative planning and patient counseling. (Level III)
Glenohumeral Arthritis
Viscosupplementation is an established nonoperative treatment for osteoarthritis of the knee. However, to date, no such therapy has proved to be successful for the treatment of shoulder pain. Blaine et al.27 evaluated the use of sodium hyaluronate (HYALGAN; sanofi-aventis, Bridgewater, New Jersey, and Fidia Farmaceutici, Abano Terme, Italy) for patients with shoulder pain resulting from glenohumeral osteoarthritis, rotator cuff tear, and/or adhesive capsulitis. Patients were randomized to receive either five weekly intra-articular injections of sodium hyaluronate, three weekly intra-articular injections of sodium hyaluronate followed by two weekly intra-articular injections of saline solution, or five weekly intra-articular injections of saline solution. When the results were stratified according to diagnosis, significant improvement was noted in patients with osteoarthritis who were managed with a three and five-injection regimen of sodium hyaluronate. No significant difference was demonstrated in patients without osteoarthritis. The safety profile was very favorable, with no serious product-related adverse effects noted. The overall findings indicate that sodium hyaluronate is effective and well tolerated for the treatment of osteoarthritis and persistent shoulder pain that is refractory to other standard nonoperative interventions. (Level I)
Patients with glenohumeral osteoarthritis often have posterior glenoid erosion. One method to address this erosion at the time of total shoulder arthroplasty is to eccentrically ream the anterior aspect of the glenoid, thereby restoring normal glenoid version. Little information exists on the degree of posterior glenoid erosion that can be corrected while still maintaining sufficient bone stock to implant a glenoid component securely. Clavert et al.28 created posterior glenoid erosions in cadaveric scapulae to simulate retroversion of =15°. The glenoid was then reshaped to correct the glenoid retroversion to neutral, and a glenoid component with central and peripheral pegs was inserted. In all experimental cases, at least one of the four pegs penetrated the glenoid vault. In one case, there was a fracture of the anterior part of the rim. The authors concluded that glenoid retroversion of =15° cannot be satisfactorily corrected simply by reaming to lower the anterior edge of the glenoid and to restore neutral version when using a glenoid component with peripheral pegs.
Outcomes of Arthroplasty
Methods to improve the accuracy of prosthetic implantation continue to evolve. Edwards et al.29 investigated the use of an image-free shoulder navigation system in a cadaver trial and in an initial group of patients undergoing shoulder arthroplasty. Before and after humeral head resection, the anatomic neck axis (retroversion and inclination) and humeral head diameter were measured with the navigation system. Native glenoid surface orientation was registered, and a navigation tracker was attached to the glenoid reamer. The navigation system recorded the change in inclination and version relative to the native glenoid during reaming. In the initial cadaver study and in the initial clinical trial, computer-aided shoulder navigation was found to be safe and to provide valuable intraoperative information. The navigation system provided real-time feedback on the humeral resection as it related to anatomic neck geometry and also provided real-time measurement of the angulation of the glenoid reamer relative to the preoperative glenoid deformity.
Hemiarthroplasty is a common treatment for cuff tear arthropathy and glenohumeral arthritis associated with a massive rotator cuff tear. However, the long-term results and preoperative factors affecting these results are not well reported. Goldberg et al.30 evaluated the functional results following hemiarthroplasty for the treatment of rotator cuff deficiency. For the group as a whole, the functional results improved significantly but function still remained limited. Of the patients who could actively elevate the arm to 90° preoperatively, 88% achieved satisfactory results according to the limited goals criteria of Neer. Patients who could actively elevate the arm preoperatively to 90° had significantly better function and pain relief and a higher total functional score than the patients who were unable to do so. The authors concluded that patients who have preoperative forward elevation of 90° benefit the most from this procedure. (Level IV)
The most appropriate treatment for advanced glenohumeral osteoarthritis remains controversial. Radnay et al.31 performed a systematic review of the literature in which total shoulder arthroplasty was compared with humeral head replacement for the treatment of glenohumeral osteoarthritis. Pain data were converted to a 100-point score. Outcome assessment data were pooled when possible, and analyses with normal test statistics were performed. Studies that provided appropriate data demonstrated that total shoulder arthroplasty provided significantly greater pain relief, forward elevation, gain in forward elevation, gain in external rotation, and patient satisfaction in comparison with humeral head replacement. The requirement for revision surgery was also significantly lower among patients managed with total shoulder arthroplasty as compared with those managed with humeral head replacement, and there was a low rate of revision (1.7%) of all-polyethylene glenoid components. On the basis of this analysis, current literature supports the use of total shoulder arthroplasty over humeral head replacement for the treatment of glenohumeral osteoarthritis. Concerns about high rates of glenoid component failure are not supported in the literature. However, glenoid component failure remains a recognized complication of total shoulder arthroplasty. This risk led Clinton et al.32 to explore humeral head replacement with glenoid reaming (the so-called ream and run technique) for the treatment of glenohumeral arthritis in active patients. The patients were evaluated preoperatively and postoperatively with use of the Simple Shoulder Test. The scores were significantly better for the total shoulder arthroplasty group as compared with the ream and run technique group for as long as eighteen months postoperatively. However, by eighteen months, the results were comparable between the two treatment groups. While many studies have highlighted the superiority of total shoulder arthroplasty over humeral head replacement, this study demonstrated that humeral head replacement combined with glenoid reaming may offer comparable functional results at the time of the three-year follow-up, although the time to recovery may be longer.
Sperling et al.33 compared the results of humeral head replacement and total shoulder arthroplasty for the treatment of rheumatoid arthritis. There was significant long-term pain relief, improvement in active abduction, and external rotation in association with both humeral head replacement and total shoulder arthroplasty. In patients with a thin or torn rotator cuff, there was no difference in improvement in terms of pain and motion when hemiarthroplasty was compared with total shoulder arthroplasty. However, in patients with an intact rotator cuff, improvement in terms of pain and abduction was significantly greater and the risk of revision was significantly lower in association with total shoulder arthroplasty. Glenoid erosion was present in 98% (sixty-two) of the sixty-three shoulders treated with humeral head replacement, whereas periprosthetic glenoid radiolucency was present in 72% (110) of the 152 treated with total shoulder arthroplasty. This study indicates that there is marked long-term relief of pain and improvement in motion in association with humeral head replacement and total shoulder arthroplasty. Among patients with an intact rotator cuff, total shoulder arthroplasty appears to be the preferred procedure for pain relief, improvement in abduction, and lower risk of revision surgery.
Complications of Shoulder Arthroplasty
Glenoid component loosening is a recognized cause of failure after total shoulder arthroplasty. However, there are few studies on the outcome of revision surgery for this problem. Cheung et al.34 evaluated the outcomes for patients who underwent revision total shoulder arthroplasty because of glenoid component loosening. One group of patients underwent reimplantation of a new glenoid component, and the other group underwent removal of the glenoid component and insertion of bone graft at the site of the deficit without glenoid reimplantation. Both groups had significant overall improvement in terms of pain. A significant improvement in active elevation occurred in the reimplantation group. The rate of survival free of reoperation at five years was 91% in the group managed with reimplantation and 78% in the group managed with component removal and bone-grafting. On the basis of this study, we can conclude that glenoid revision surgery often will lead to pain relief and patient satisfaction. Additionally, there appears to be a slight clinical benefit to reimplanting a glenoid component whenever doing so is structurally possible. (Level IV)
When reimplantation is not immediately possible, component removal and bone-grafting is performed in an effort to reconstitute bone stock for possible later reimplantation. Cheung et al.35 evaluated a group of patients who had undergone glenoid component removal with bone-grafting of the defect and who later underwent glenoid component revision. Two patients had positive intraoperative cultures and had continued pain requiring repeat revision surgery. The remaining patients expressed satisfaction with the procedure and stated that they felt better following the surgery. The authors demonstrated that reimplantation of a glenoid component into a previously grafted bed provides pain relief for most patients; however, motion was not reliably improved. (Level IV)
Infection is a devastating complication of total shoulder arthroplasty. Currently, two-stage reimplantation is the accepted method for the treatment of this problem. However, there is little information on the outcome of two-stage reimplantation. Strickland et al.36 reported the results for nineteen patients who underwent a two-stage reimplantation for the treatment of a deep infection after shoulder arthroplasty. In that study, there were two excellent results, four satisfactory results, and thirteen unsatisfactory results. The infection was considered to be eradicated in only 63% of the patients. Improvements in terms of pain and range of motion were modest. This study suggests that two-stage reimplantation for the treatment of infection at the site of a shoulder arthroplasty is associated with a high rate of unsatisfactory results, marginal success in terms of eradicating infection, and a high complication rate.
Glenohumeral arthrodesis performed after the failure of a shoulder arthroplasty is considered a salvage procedure when all other reconstructive options have failed. Scalise and Iannotti37 performed a retrospective review of the outcomes of glenohumeral arthrodesis for the treatment of severe glenohumeral bone loss and deltoid muscle and rotator cuff insufficiency. Five of the seven patients demonstrated an intact fusion at the time of follow-up, with four of the seven patients requiring additional bone-grafting procedures in an effort to obtain union. Overall, the average clinical outcome score improved significantly. The most common complication was delayed union requiring additional surgical procedures. The results of this study suggest that glenohumeral arthrodesis can yield satisfactory clinical outcomes. However, both the patient and the surgeon should be aware of the complex nature of this procedure and the frequent need for additional procedures to obtain fusion. (Level IV)
Reverse Total Shoulder Arthroplasty
Reverse total shoulder arthroplasty provides an alternative to standard total shoulder arthroplasty for the treatment of selected complex shoulder problems. The growth in the clinical application of this technique has allowed an evaluation of the results according to etiology. Wall et al.38 grouped patients according to etiology, and the clinical and radiographic outcomes for each group were measured. Although substantial clinical and functional improvement was observed in all groups, patients with primary rotator cuff tear arthropathy, primary osteoarthritis with a rotator cuff tear, and a massive rotator cuff tear had better outcomes, on the average, than did patients who had posttraumatic arthritis and those undergoing a revision arthroplasty. The advanced age of the patients in their series and the relatively short duration of follow-up suggest that the prosthesis should be used with caution. (Level II)
Fractures
Proximal Humeral Fractures
The length of immobilization when treating proximal humeral fractures nonoperatively is not well established. In the study by Lefevre-Colau et al.39, the results of treatment with early mobilization of the shoulder were compared with those of conventional treatment (three weeks of immobilization followed by physical therapy). Patients with impacted proximal humeral fractures were randomly assigned to receive early passive mobilization or conventional treatment. At three months and at six weeks, the early mobilization group had significantly better Constant scores and had better active mobility in forward elevation in comparison with the conventional treatment group. At three months, the early mobilization group had significantly reduced pain in comparison with the conventional treatment group. No cases of fracture displacement or nonunion were noted. The authors demonstrated that early mobilization for the treatment of impacted proximal humeral fractures is safe and more effective for quickly restoring the physical capability and performance of the injured arm than is conventional immobilization followed by physiotherapy. (Level I)
Plate fixation of displaced proximal humeral fractures has increased with the introduction of proximal humeral locking plate technology. Owsley and Gorczyca40 assessed the radiographic and clinical results for a group of patients who were managed with a proximal humeral locking plate and a standard postoperative rehabilitation protocol for the treatment of a displaced proximal humeral fracture. Fracture-healing was noted in 98% of the patients at six months. However, complications were noted radiographically in 36% of the patients; the complications included screw cutout with intra-articular displacement (23%), >10° of varus displacement (25%), and osteonecrosis (4%). These radiographic findings were noted in 57% of patients more than sixty years of age and in 22% of those less than sixty years of age. This age-based difference was significant. Screw cutout occurred in 43% of the patients older than sixty years of age. Patients with a complication had significantly worse functional outcomes. These results indicate that the use of locking plates for the treatment of proximal humeral fractures is associated with a high rate of screw cutout, especially in patients older than sixty years, resulting in compromised function. (Level IV)
The rate of pulmonary embolism after lower extremity and pelvic trauma is well known. However, there is limited information on the risk of pulmonary embolism after upper extremity surgery. Hoxie et al.41 performed a retrospective review of patients who underwent operative treatment of acute, isolated proximal humeral fractures over a five-year period to determine the risk of pulmonary embolism. Postoperatively, seven patients sustained a pulmonary embolism that was confirmed with computed tomography. Of these, four patients had undergone a hemiarthroplasty and three had undergone open reduction and internal fixation of the fracture. The overall incidence of pulmonary embolism in this series was 5.1%; none of the emboli were fatal. This study raises the question of whether prophylactic anticoagulation is needed after routine proximal humeral fracture repair.
Clavicular Fractures
Nonunion or malunion of midshaft clavicular fractures is a potential cause of disability with local pain or neurovascular impairment. Rosenberg et al.42 evaluated the functional outcome for patients managed with open reduction and internal fixation with autogenous bone graft. After an average duration of follow-up of forty-one months, osseous union was achieved in all patients. The average Constant score for the affected shoulders remained significantly lower than that for the normal, contralateral shoulders at the time of follow-up. Ten patients reported various degrees of pain, and only three patients were pain-free. Only 46% of the patients returned to their previous professional and recreational activities. The authors demonstrated that, although solid union is consistently achieved with realignment and fixation of a symptomatic nonunion or malunion of the middle part of the shaft of the clavicle, patients can remain functionally impaired.