16.1 Introduction

Originally developed to treat the rotator cuff-deficient shoulder, reverse shoulder arthroplasty (RSA) is a relatively new technique in shoulder surgery. Interestingly, the first shoulder replacement performed by Pean in 1891 for tuberculosis was an RSA. The reverse has evolved, like all technology, in fits and spurts into a novel technique with which we can now treat the arthritic rotator cuff-deficient shoulder and other complex shoulder problems that previously were only marginally amenable to surgical reconstruction.

Surgical indications for RSA have significantly expanded to include a variety of complex pathologies. Additionally, the surgical technique, implants, and instrumentation for RSA have all evolved significantly since this prosthesis was first implanted—an evolution that continues to this day. The essence of this evolution can be better understood through an understanding of clinical outcomes for patients undergoing RSA. RSA revolutionized the treatment of the cuff-deficient shoulder, and provided a vast improvement in the clinical outcomes from the historic treatments available. Nevertheless, initially, surgical indications for RSA were narrow, and results following RSA were unpredictable, with high rates of complications. As surgeons better understood the etiology of complications and how to avoid them, indications for RSA were refined and results became more predictable. Today, armed with a better understanding of the bio-mechanics of RSA, surgeons are expanding the indications for RSA, and short- and long-term results following RSA are improving even further. Many issues regarding RSA have been resolved, but as these questions are answered, more are being asked.

16.2 Historical Perspectives

The painful rotator cuff-deficient shoulder has long been a challenging condition to treat, and prior to the development of RSA, results of the surgical treatment of these patients were abysmal. By the late 1960s, the advent and success of modern hip replacement using polyethelene cemented with methyl methacrylate mated with a cobalt chrome spherical femoral head led surgeons to successfully expand their indications to the shoulder. However, specific subsets of patients remained challenging to treat. Many of these patients had rheumatoid arthritis, with the accompanying deficient rotator cuff and soft-tissue envelope. Other patients did not have RA, but did have advanced glenohumeral degenerative joint disease and rotator cuff deficiency. These soft-tissue deficiencies and degenerative changes, often superimposed upon a paucity of scapular bone, presented a unique challenge to the existing technology.

In the seminal paper on cuff tear arthropathy, Neer et al recognized that “surgical reconstruction of these shoulders is especially difficult.”1 The unique difficulty of treating these shoulders is due to the nonfunctional rotator cuff; the irreparable rotator cuff tear leads to vertical instability of the glenohumeral joint, in turn leading to glenohumeral degenerative joint disease, and pain from both the glenohumeral and acromiohumeral articulations. The reconstructive options to date all had specific shortcomings, and failed to create a blend of joint stability, function, and prosthetic durability.2 Neer, among others, had attempted such salvage procedures as radical acromionectomy3 and glenohumeral fusion; the morbidity of radical acromionectomy is, with hindsight, self-evident, while glenohumeral fusion provided pain relief but restriction in motion and function in extremis.1,4 Hemiarthroplasty had shown modest improvements in pain, but did not improve motion and often lead to progressive pain due to glenoid erosion. Anatomic total shoulder arthroplasty had the dual problem of poor functional improvement and eventual pain from loosening of the glenoid component.1,5,6

In an effort to address these issues, a number of surgeons began reversing the components by placing the sphere on the scapula and the socket onto the resected humerus. The experience of early shoulder pioneers, Dr. Lipmann Kessel, reflected the challenges of surgical reconstruction of these patients. Constrained prostheses and early reverse ball and socket designs were developed as alternatives to standard arthroplasty designs, but the early experiences were characterized by disastrous results due to catastrophic loosening of the glenoid component.7,8,9,10,11,12,13,14,15,16 The etiology of failure of these early reverse prostheses was due to the biomechanics of the implants. The center of rotation was lateral to the implant-glenoid interface. This lateralization leads to significant shear stresses, implant loosening, and ultimately catastrophic failure.17 However, hidden in the early experience with reversed prostheses were principles that would eventually lead to success. In an article in 1979 by Kessel in a period of 3 years, 13 patients (14 shoulders) were satisfied and would have the operation again, and 11 (12 shoulders) patients were enthusiastic. The author’s criteria were more stringent. In terms of composite grading, there was, when all factors were considered, an overall success rate of 72% reported using this reversed semicaptive prosthesis.

Working in parallel with Kessel, Dr. Paul Grammont, a French surgeon, developed a reverse ball and socket prosthesis in 1985 that was the first to successfully address the biomechanical obstacles that so many other surgeons had been unable to overcome; with the development and release of his prosthesis, Grammont revolutionized surgery of the shoulder. Grammont was a thoughtful surgeon, as well as a student of comparative and evolutionary anatomy. Grammont hypothesized that as humans evolved from quadrupeds to walking upright, the shoulder was left with certain anatomic and biomechanical disadvantages, predisposing the rotator cuff to fail and, in turn, the shoulder as a whole to lose function. This evolutionary anatomic disadvantage helped Grammont to dispense with the in essence anatomic solutions for cuff-deficient degenerative shoulders that had been proposed to date and pursue a nonanatomic but nevertheless functional solution.2 To this end, Grammont’s first prosthesis consisted of a polyethylene humeral component and metallic glenoid component, or glenosphere. The early experience with this prosthesis taught Grammont some useful lessons, allowing him to design the Delta III prosthesis; this is the first implant that would be recognized today as a “reverse” and is still being commonly implanted today, a testament to the sound biomechanical principles on which the prosthesis was designed. The biomechanical principles that distinguished the Delta III from other reverse ball prostheses, and subsequently made the implant more durable, are as follows: (1) the glenosphere is not significantly lateralized, placing the center of rotation directly at the bone-prosthesis interface, minimizing shear stresses on the implant; (2) placing the glenosphere more inferiorly effectively lengthens the humerus, thus tensioning the deltoid and allowing the deltoid to function in place of the deficient rotator cuff; (3) the glenosphere was increased in size relative to other contemporary designs, allowing a greater range of motion with improved stability.17

16.3 Early Clinical Results

The early results of Grammont’s prosthesis were modestly successful by modern standards, but represented a quantum leap in improvement relative to other reverse ball and socket designs. In his first published report on his experience with reverse prosthesis, Grammont reported his results in eight patients. All patients reported excellent pain relief, but only half of the patients were able to elevate their arm beyond 100 degrees. However, the poor functional results were the result, in most cases, of a nonunion of an acromial osteotomy, Grammont’s preferred surgical approach.17 Although only half of the patients in this series had an acceptable functional result, this was a clear improvement and an inspiration for Grammont to improve upon and refine his prosthetic design.

Grammont’s Delta III prosthesis was the first reverse ball and socket prosthesis to be adopted in widespread use with acceptable results.17,18,19,20,21 The early experience with this prosthesis was predominantly in Europe, specifically in France, as the reverse prosthesis did not receive Food and Drug Administration (FDA) approval in the United States until 2003. Although surgeons utilized the Grammont’s prosthesis for a small number of revision, fracture malunion, and other indications, a majority of the early cases were for rotator cuff arthropathy. The early experience of these surgeons was variable, but definite trends emerged with an analysis of these results (► Table 16.1). First of all, the Grammont’s prosthesis reliably relieved pain, and improved objective outcome and functional scores. Additionally, the Delta III predictably increased forward elevation of the arm beyond 100 degrees, while improvements in external and internal rotation were variable, if present at all.