➢ Shoulder stiffness affects a diverse population of patients suffering a decrease in function and shoulder pain. Arthroscopic management of this debilitating spectrum of pathology is a safe and effective course of action in cases recalcitrant to nonoperative therapy.
➢ Arthroscopic management of the stiff shoulder has been reported to be effective in the treatment of stiffness due to adhesive capsulitis, birth palsy, stiffness in the setting of rotator cuff tears, and osteoarthritis in the posttraumatic patient, in the postoperative patient, and in the throwing athlete.
➢ Arthroscopic management is most effective in treating the stiff shoulder in the setting of adhesive capsulitis recalcitrant to nonoperative therapy or posttraumatic stiffness. Results are more guarded in the treatment of postoperative stiffness.
➢ Excessive force and trauma to the shoulder, including fracture of the humerus, can be avoided with a 360° capsular release for shoulder stiffness rather than manipulation under anesthesia.
Shoulder stiffness is a common and poorly understood cause of pain and dysfunction across a diverse patient population. This condition is synonymous with a decreased passive range of motion that may limit an individual’s ability to work, to perform activities of daily living, and to enjoy leisure activities1-3. Pathologic loss of shoulder range of motion was first described by Codman4; then, later, Nevaiser coined the term “adhesive capsulitis.”5 When the onset of stiffness is idiopathic in nature, it is considered primary adhesive capsulitis. If stiffness is related to a preceding injury or other cause, it is considered secondary adhesive capsulitis. Prevalence of primary adhesive capsulitis is estimated at 2.4 cases (95% confidence interval, 1.9 to 2.9 cases) per 10006. The incidence of frozen shoulder in the general population is <2%. The incidence in some populations such as those with diabetes mellitus is approximately 10% in patients with type-1 diabetes and up to 29% in patients with type-2 diabetes7,8.
The incidence of secondary adhesive capsulitis is difficult to quantify as this occurs as a result of trauma, a surgical procedure, degenerative disease, birth palsy, or overuse or disuse of the shoulder.
Treatment of shoulder stiffness, regardless of cause or etiology, should address the underlying anatomic cause and inciting pathologic process. Physical therapy, rest, and anti-inflammatory medications locally or systemically are the current mainstay of treatment1,9,10. However, should these therapies fail, operative intervention is warranted10. Prior to advancements in arthroscopy, the treatment for recalcitrant cases of stiffness was manipulation under anesthesia11-14. Manipulation is successful in 94% to 100% of cases, with a small complication rate11,12. Complications include intraoperative humeral fractures, glenoid fractures, and rotator cuff tears1. Arthroscopy for lysis of adhesions is an alternative to manipulation under anesthesia alone and adds the benefit of intra-articular examination and confirmation of the diagnosis without the risks of manipulation. Outcomes described for manipulation under anesthesia show that, even in a fifteen-year follow-up, remanipulation and complication rates are small1. In addition, a recent systematic review revealed that arthroscopic capsular release yielded better postoperative abduction and external rotation when compared with manipulation under anesthesia1. However, there was no significant difference in patient-reported outcomes (American Shoulder and Elbow Surgeons Assessment Form15, Constant Shoulder Score16, Simple Shoulder Test17, and visual analog scale for pain) across groups. Current literature, as a whole, demonstrates arthroscopic treatment of the stiff shoulder to be as effective as, if not more effective than, manipulation10,18,19. The use of arthroscopic management of stiffness has grown beyond the treatment of simple primary adhesive capsulitis to treatment of stiffness from secondary causes, including rotator cuff tears, osteoarthritis, trauma, sequelae of arthroplasty, and birth palsy. The aim of this article is to review the pathologic processes that lead to a stiff shoulder and their treatment with arthroscopic capsular release.
Primary Adhesive Capsulitis
Pathoanatomy and Treatment Considerations
Primary adhesive capsulitis begins as an inflammatory process that evolves into a fibrotic reaction. It has classically been divided into phases of freezing (insidious onset of painful and progressive loss of range of motion), frozen (subsidence of pain and plateauing stiffness), and thawing (gradual improvement and resolution of symptoms)20. Initiators of this pathoanatomy are poorly understood and there are many unknowns in the treatment of this ubiquitous disease. The mainstay of initial therapy is conservative, including targeting inflammation and pain with the use of anti-inflammatory medications and corticosteroids21. Physical therapy also plays a role as an initial intervention, but results are mixed as to its utility21,22.
The senior author’s preferred treatment approach consists of initial nonoperative therapy in the form of oral nonsteroidal anti-inflammatory drugs (NSAIDs), intra-articular injections or corticosteroids, and home therapy. Should this regimen prove to be ineffective for six months after the initial presentation of symptoms, the patients are counseled on the natural history of primary adhesive capsulitis and an informed decision is made either to continue conservative treatment for the ten to fourteen months needed to reach the thawing phase or to treat with arthroscopic capsular release.
History of Arthroscopic Capsular Release in Primary Adhesive Capsulitis
Segmüller et al. first described the use of arthroscopic capsular release for the treatment of primary adhesive capsulitis in 199523. These investigators followed a cohort of twenty-four patients for whom nonoperative treatment failed and who later underwent arthroscopic capsular release. At one year postoperatively, 88% of these patients were satisfied with their outcome, without intraoperative or postoperative complication. The technique utilized in this and other early descriptions divides the inferior capsule to the three o’clock position anteriorly and the nine o’clock position posteriorly. Later techniques described the circumferential release of the capsule, with favorable results and only one patient with a complication (transient axillary nerve neurapraxia)24-26. Technical refinements, including innovative instrumentation and extra-articular dissection, after intra-articular access have made a 360° capsular release a safe and reliable procedure for the experienced arthroscopist7,27. Video 1 demonstrates extra-articular and intra-articular dissection of the axillary pouch to complete the inferior capsular release and axillary nerve decompression in a fifty-five-year-old patient with primary adhesive capsulitis recalcitrant to seven months of nonoperative therapy. Lafosse et al. described a 360° release technique in which the rotator interval is first released from a subacromial (extra-capsular) midlateral portal. This grants anterior access to the joint, allows for extra-capsular release about the subscapularis tendon, and removes the primary restriction to external rotation27. The dissection is then taken down the anterior capsule as inferiorly as possible. With the addition of an anterior portal as well as a posterior portal, a complete capsular release is performed27. In severe cases, extra-capsular structures such as the subscapularis tendon become adherent and are a further cause of stiffness. In that case, Liem et al. proposed the intra-articular release of the subscapularis tendon28. This has been shown to improve range of motion without leading to instability as a consequence of release28.
Outcomes of Arthroscopic Capsular Release in Primary Adhesive Capsulitis
Intermediate-term (two to ten years postoperatively) results of arthroscopic capsular release, in primary and secondary adhesive capsulitis, have demonstrated a significant difference in postoperative outcome measures compared with preoperative outcome measures, including range of motion and patient satisfaction, across a cohort of forty-three patients at a mean follow-up time of seven years (p < 0.001)19,29. Long-term outcomes, as long as seven years, following index arthroscopy show that not only do patients do well, but they also continue to improve as years pass19. Le Lievre and Murrell demonstrated a decrease in patient-reported pain and no deficits in motion compared with the contralateral shoulder at one, six, twelve, and fifty-two weeks and finally seven years after the surgical procedure for primary adhesive capsulitis19. In patients with diabetes, the literature is mixed on whether arthroscopic release provides effective and lasting benefits2,8,19. Patients with diabetes show significant improvement (p < 0.01 in a study of twenty-one patients per group) when comparing preoperative and postoperative outcomes; however, they tend to fare slightly worse than their non-diabetic counterparts in the majority of comparison studies (which showed either significance or a trend that did not reach significance)8,30.
The comparison between arthroscopic capsular release and manipulation under anesthesia began as early as 1995 with Ogilvie-Harris et al. comparing two similar groups of patients, one treated with manipulation under anesthesia and diagnostic arthroscopy and the other treated with arthroscopic capsular release alone2. Both groups fared well postoperatively, with the capsular release group showing a significant decrease (p < 0.01) in pain and increased range of motion compared with the manipulation-under-anesthesia group. A recent systematic review corroborated these findings and demonstrated that patients treated with either method do well, but better improvements in range of motion were seen in patients who underwent capsular release1. However, the review included twenty-one Level-IV studies that followed a total of 989 patients over thirty-five months and showed no significant difference in patient-reported outcomes between patients treated with arthroscopic capsular release and those treated with manipulation under anesthesia1.
Secondary Causes of Stiff Shoulder
The underlying pathophysiology of shoulder stiffness in patients who sustain a painful decreased range of motion secondary to rotator cuff tears, osteoarthritis, trauma, arthroplasty, or birth palsy remains similar to that of primary adhesive capsulitis. However, with regard to the stiff shoulder due to secondary causes, it is important to note that treatment also remains largely similar to that of primary adhesive capsulitis, with nonoperative modalities as the mainstay and operative intervention as an option in recalcitrant cases. A comparison of outcomes of the arthroscopic treatment between primary and secondary adhesive capsulitis demonstrated significant improvement (p < 0.05) in both patient populations3.
Shoulder Stiffness and Concomitant Rotator Cuff Tear
It is well known that patients who present with a rotator cuff tear often also present with shoulder stiffness. Some amount of stiffness is present in as high as 39% (twenty-eight of seventy-two) of cases31. The mainstay of treatment for such patients has been physical therapy and nonoperative interventions to relieve stiffness prior to rotator cuff repair31,32. This is a difficult proposition for some patients as rehabilitation causes pain and a propensity toward noncompliance. Thus, several studies have investigated the treatment of concomitant rotator cuff tears and shoulder stiffness with a combination procedure including arthroscopic rotator cuff repair as well as capsular release31,33,34. Results at one year postoperatively showed that patients diagnosed with a moderately stiff shoulder as well as a rotator cuff tear performed just as favorably as patients with a rotator cuff tear alone33,34. Caveats to this include that patients with stiffness require a longer time to achieve postoperative return of function and have decreased strength and that patients with severe preoperative stiffness have poor outcomes after rotator cuff repair31,33.
Postoperative Stiffness After Rotator Cuff Tear
The incidence of postoperative stiffness after rotator cuff repair has also been described in several studies31,35,36. Of the 795 patients in the study by Denard et al., the incidence of transient stiffness responsive to nonoperative treatment was 10% (seventy-nine patients) and resistant stiffness that required operative intervention or that was permanent was 3.3% (twenty-six patients)35. In a study by Parsons et al., patients were immobilized after rotator cuff repair for a period of six weeks, after which 23% (ten of forty-three patients) were found to have a decreased passive range of motion and were deemed “stiff”37. The remainder of patients had near-physiologic range of motion at that time and became the “nonstiff” cohort. Following the patients for one year without intervention demonstrated that the patients who were initially stiff not only regained their range of motion but showed better healing rates on magnetic resonance scans when compared with the nonstiff group. When patients fail to regain range of motion, postoperative stiffness treated with arthroscopic capsular release has shown promise in several studies24,36,38. Patients’ overall range of motion improved and patients were satisfied with the procedure36.
Shoulder Stiffness in Osteoarthritis
The goal of operative intervention in patients with osteoarthritis of the shoulder includes decreasing pain and increasing function. Increasing function in these patients includes overcoming the stiffness that often accompanies the fibrosis and bone formation present as part of the disease process39. Although it is well known that total shoulder arthroplasty is a reliable surgical option for the treatment of osteoarthritis of the shoulder, patients often request joint-sparing alternatives. In addition, patients who are younger than fifty years of age and undergo hemiarthroplasty or total shoulder arthroplasty often have concerning radiographic findings within seven to ten years of the surgical procedure, including periprosthetic lucency and humeral subluxation40,41. Arthroscopy alone has been used for the treatment of osteoarthritis, with favorable short-term results42,43. Several authors have described arthroscopic debridement with the addition of capsular release and axillary nerve neurolysis39,44. However, overall supportive literature for arthroscopic treatment of osteoarthritis is lacking45. Although results are mixed, patients who are young and active are poor candidates for total shoulder arthroplasty and thus warrant an attempt at joint preservation45. Figure 1 demonstrates a fifty-year-old male patient with glenohumeral arthritis and stiffness who was treated successfully with arthroscopic capsular release and osteophyte removal.
Although only small cohorts exist, reported outcomes of patients who underwent arthroscopic capsular release for a stiff shoulder after a fracture have yielded favorable results46-48. Holloway et al. designed and executed one such study in which they compared three groups of patients with stiff shoulder: a postoperative cohort, a post-fracture cohort, and, finally, an idiopathic capsulitis cohort46. They found that all three groups showed improvements in pain and function but that patients in the fracture group and the idiopathic group fared better than those in the postoperative cohort. In a study of twenty-one patients over one year, Levy et al. reported a significant improvement (p < 0.05) of posttraumatic stiffness recalcitrant to conservative therapy after arthroscopic capsular release48. Patients were followed over one year, during which time their immediate postoperative range of motion was excellent; the range of motion decreased by 48% at six months but then steadily improved to 110% of the immediate postoperative range of motion by the culmination of the study. Figure 2 demonstrates a clinical vignette from our institution of a twelve-year-old female patient involved in a motor vehicle accident who sustained a proximal humeral fracture and was treated arthroscopically one year later for stiffness.
Stiffness in Shoulder Reconstruction
In patients who require hemiarthroplasty for fracture, therapy generally begins with passive range of motion starting on the first postoperative day49. In an attempt to improve the chances for tuberosity healing, Barth and Burkhart described an approach whereby hemiarthroplasty follows six weeks of immobilization, prioritizing bone-healing and integration over the prevention of stiffness50. They then performed an arthroscopic release of the capsule after bone-healing if the patient had a deficient range of motion or pain with passive stretch. Although only a few illustrative cases have been presented to our knowledge, patients later have radiographic healing of the tuberosities as well as near full range of motion after hemiarthroplasty and subsequent arthroscopic release50. This is not the generally accepted practice in the treatment of proximal humeral fractures with hemiarthroplasty, but it can be implemented when there is heightened concern for fracture-healing due to bone quality and predisposing factors to malunion (i.e., age, smoking status, and fracture pattern).
Brachial plexus birth palsy occurs at a rate of about four per 1000 births in the United States51. Brachial plexus palsy persists into childhood at a rate of 0.4 per 1000 births. External rotation deficit leads to internal rotation contractures and may cause glenohumeral deformity. The method by which internal rotation contractures are treated remains controversial. An open, anterior approach to the shoulder is the historical and, in some institutions, current method of treatment52. Gilbert et al. described open release in addition to release and reflection of the subscapularis tendon, with the addition of latissimus tendon transfer in patients older than four years of age53-55. Concerns about efficacy and cosmesis, as well as the advances in arthroscopic shoulder surgical procedures in adults, led to the consideration of arthroscopic release of internal rotation contracture in children. This technique, with or without subscapularis tendon release, has been proven to be effective in regaining at least 45° of passive external rotation in most described cases56-58. In a series of forty-one patients by Pearl, all but one patient, who had a severe contracture with concomitant glenohumeral deformity, were treated successfully with arthroscopy alone56. A similar study was repeated at a longer follow-up with the addition of magnetic resonance imaging (MRI) and showed that arthroscopic release in patients who had a mean age of four years yielded diminished retroversion of the glenoid and improved external rotation and eleven of eighteen children had active external rotation five years postoperatively59. In that study, Breton et al. suggested that contractures, as a sequela of birth palsy, do not necessarily require tendon transfer for glenoid remodeling and active external rotation and can be treated arthroscopically59.
Posterior capsular contracture results in overhead-throwing athletes as a consequence of serial insults to the posteroinferior capsule in the deceleration phase of throwing60,61. Arthroscopic management consists of identification and release of the posterior band of the inferior glenohumeral ligament along with the posteroinferior capsule62. Good results have been demonstrated for patients who present with isolated loss of internal rotation as well as patients with concomitant labral pathology, albeit in small cohort studies. In these studies, patients regain internal rotation and have a reduction in pain postoperatively. In a study of sixteen patients, Yoneda et al. found that eleven patient athletes were able to return to their previous level of play and throwing pain completely disappeared in fourteen patients63.
Effectiveness of Arthroscopic Treatment of Stiffness Based on Etiology
In a retrospective study of patients who underwent arthroscopic capsular release for shoulder stiffness after failure of conservative treatment, Elhassan et al. compared the effectiveness across specific etiologies10. The study compared posttraumatic, postoperative, and idiopathic etiologies for shoulder stiffness. At a mean time of forty-six months postoperatively, they found a significant decrease (p < 0.0001) in pain as well as an increase in function across all groups in a cohort of 115 patients. They also showed that there was no difference in outcomes between the idiopathic and posttraumatic groups, which both performed better than the postoperative group of patients.
Arthroscopic capsular release can result in complications, albeit rarely. In a retrospective study with a minimum five-year follow-up, Le Lievre and Murrell reported no intraoperative complications in addition to no long-term complications, including osteoarthritis, recurrence requiring reoperation, axillary nerve dysfunction, infection, or shoulder instability19. Comparing manipulation under anesthesia and capsular release, Amir-Us-Saqlain et al. found the incidence of serious complications due to manipulation under anesthesia to be 0.4% (proximal humeral fracture) and that due to capsular release to be 0.6% (transient axillary neurapraxia)64. Although the difference in complication rate is negligible, all of the complications of manipulation under anesthesia were proximal humeral fractures. In comparison, Grant et al. found that the complication from capsular release most commonly involved superficial wound infection and secondarily involved transient axillary nerve neurapraxia1.
Despite the 360° release of the capsule as well as all of the glenohumeral ligaments, instability is not a major concern after arthroscopic release of the stiff shoulder. In a review of English-language literature, one instance of subjective postoperative instability65 and one case report of dislocation following arthroscopic release are reported66. In addition, the capsule and synovial lining reconstitute following release, thus restoring the fluid seal of the shoulder to maintain the nourishing effects of synovial fluid to intra-articular structures.
Our Preferred Technique
The senior author’s preferred treatment algorithm for the stiff shoulder, regardless of etiology, almost always begins with nonoperative management. If the patient has undergone nonoperative treatment in the form of oral nonsteroidal anti-inflammatory drugs, intra-articular corticosteroid injection, and a home therapy program for a period of three to six months, arthroscopic capsular release is discussed.
Our preferred technique is a modification of the technique described by Harryman et al.26 with the addition of axillary nerve decompression as well as the use of a three-portal (two posterior and one anterior) technique rather than a four-portal technique as described by Harryman et al.
Stiffness of the shoulder affects patients of all ages and causes functional limitations for myriad reasons. Results of the use of arthroscopic capsular release are encouraging for many specific etiologies that result in a stiff shoulder. As arthroscopists continue to refine the technique and high-quality evidence of its efficacy begins to emerge, arthroscopic capsular release is proving to be a versatile tool for the treatment of the stiff shoulder.
Investigation performed at the Department of Orthopaedic Surgery, Duke University Medical Center, Durham, North Carolina
Disclosure: None of the authors received payments or services, either directly or indirectly (i.e., via his or her institution), from a third party in support of any aspect of this work. One or more of the authors, or his or her institution, has had a financial relationship, in the thirty-six months prior to submission of this work, with an entity in the biomedical arena that could be perceived to influence or have the potential to influence what is written in this work. No author has had any other relationships, or has engaged in any other activities, that could be perceived to influence or have the potential to influence what is written in this work. The complete Disclosures of Potential Conflicts of Interest submitted by authors are always provided with the online version of the article.
- Copyright © 2016 by The Journal of Bone and Joint Surgery, Incorporated