The most common long-term complication following TSA is glenoid loosening, which accounts for approximately 24% of all complications. A recent systematic review reported that asymptomatic radiolucent lines occurred at a rate of 7.3% per year after primary ATSA, with symptomatic glenoid loosening and surgical revision occurring at rates of 1.2% and 0.8% annually, respectively. Radiolucent lines around glenoid components are common; however, they do not necessarily precede symptomatic loosening or indicate a need for revision (FIGURE 33.4). The etiology of glenoid loosening is likely multifactorial and may be related to implant design, surgical technique, patient characteristics, and the integrity of the rotator cuff. Knowledge of the native glenoid anatomy and pathology, indications and techniques for implantation, mechanisms of failure, and the rationale behind various implant designs allows the surgeon to minimize complications and maximize outcomes following ATSA. Franklin et al developed a system for classifying radiolucency around keeled glenoid components that was later adapted by Lazarus et al to classify loosening around pegged components (FIGURE 33.5). These classification systems are frequently utilized and have become the standard; however, the interobserver and intraobserver reliability of each has not been established.⁵

The classic mechanism by which anatomic glenoid components loosen over time is the “rocking-horse” phenomenon (FIGURE 33.6). When the component is edge-loaded, it is compressed at one side, resulting in tensile forces on the opposite side. The resulting micromotion or rocking eventually results in compromise
at the bone-cement interface. Although this theory has not been proven conclusively in clinical or laboratory studies, it is supported by clinical experience. Edge-loading can be worsened by glenohumeral instability and rotator cuff dysfunction, ultimately leading to a rapid progression of glenoid loosening. Similar edge-loading effects and excessive implant micromotion are thought to occur when glenoid components are implanted in retroversion or superior inclination. Eccentric implant wear patterns and implant micromotion likely exacerbate the process of loosening through generation of polyethylene wear debris and particle-induced osteolysis. Another factor that contributes to loosening is potential compromise of the bone-cement interface as a result of thermal necrosis during the cementing process. Stress-shielding of bone adjacent to a rigid cement mantle or adjacent to an uncemented metal-backed component has also been implicated as a cause of radiolucent lines and clinical loosening. The influence of inflammation and biologic factors is likely of great importance, although it remains poorly understood. Inflammatory cytokines can be part of the arthritic disease process leading to shoulder arthritis and can also be increased as a response to foreign body wear particles from metal or polyethylene. These cytokines may share a role in the development of glenoid component loosening over time.⁵

Efforts to decrease glenoid loosening have resulted in changes to prosthetic design and implantation techniques. Currently, a wide variety of glenoid component options are available, including metal-backed, all-polyethylene, hybrid, bone ingrowth or ongrowth, inset, and augmented designs (FIGURE 33.7). Many recent clinical and biomechanical studies have examined these implant options. A thorough knowledge of glenoid anatomy, pathology, implant options, indications, and principles of implantation is necessary to optimize the outcome following ATSA and ultimately, to decrease the incidence of glenoid component loosening.⁵

Modern glenoid components are available in several shapes and sizes. Some implants are pear-shaped to mimic the shape of the normal glenoid, whereas others are elliptical. An anatomic pear shape offers the potential for less implant overhang superiorly and less uncovered bone inferiorly, but this shape has not been shown to be superior to elliptical designs. This may be because the arthritic glenoid is not often pear-shaped, and properly
sized elliptical implants often fit well after reaming. The backside of the components may be flat or curved. Curved designs have the potential advantage of resisting micromotion more effectively than flat designs. Curved implants theoretically convert shear stresses to compressive stresses to improve the stability of the implant. In a radiographic comparison of flat and convex components, Szabo et al reported that, at 2 years, glenoid designs with
a curved back had better seating and significantly better radiolucency scores than flat back components. In a follow-up study, however, the same patients showed no difference in progression of radiolucent lines at 10 years. Recently, inset glenoid designs have been developed to aid implant stability. By maintaining a peripheral rim of bone, displacement and micromotion can be minimized by preventing edge-loading.⁵ It provides a partial resurfacing while preserving the glenoid rim, capsule, and labrum. To date, only limited data are available to document the outcomes with this component design. Davis reported seven patients who underwent TSA with an inset glenoid component for severe glenoid bone loss. At a mean follow-up of 34 months, there were significant improvements in range of motion, pain, and Single Assessment Numeric Evaluation scores with no complications or revisions. Gunther and Lynch reported on seven patients who underwent TSA with an inset glenoid component for severe glenoid bone stock deficiency with an average 4 years of follow-up. All radiographs were classified as “low risk” for glenoid loosening. Although more research is necessary, this may be a future option in severe glenoid deformity and deficiency.⁷ The development of augmented glenoid implants has introduced another design parameter that may impact the risk of loosening.⁵ A finite element analysis comparing two different all-polyethylene anatomic glenoid designs showed that a wedge-type of augmented glenoid provided better implant fixation and stress profiles with less micromotion than a step-type design. Longer-term clinical outcome studies are needed to determine the impact of these design modifications on radiolucencies and loosening (FIGURE 33.8).⁸

The ideal conformity between the humeral and glenoid implant has not yet been determined. A fully conformed articulation, such as the original Neer prosthesis, may uniformly distribute stress at the implant-bone interface (FIGURE 33.9). However, normal glenohumeral translation can then occur only with articular separation and edge-loading. Translation can occur more freely with less conformity between the implants, but then contact pressures are not uniform.⁵ A biomechanical study looking at different radial mismatches (RMs) demonstrated that greater than 10 mm of RM resulted in loosening of the glenoid component.⁹ Another biomechanical and finite element analysis showed that greater RM has the advantage of providing greater glenohumeral stability but with higher implant and cement mantle stress levels and micromotion, which was worse when using a step-cut than a full-wedge design for augmented glenoids.¹⁰

Walch et al reported that there were fewer radiolucent lines at 2-year follow-up when the RM was at least 6 mm. However, they did not establish an upper limit for the mismatch. Schoch et al evaluated 451 TSAs at a mean follow-up of 5.4 years using a variation of RM between 3.4 and 7.7 mm and showed no statistically significant difference between the groups with respect to the incidence of glenoid radiolucent lines or Lazarus
score. This finding suggests that optimal RM may extend below 6.0 mm, as previously recommended by Walch et al, without affecting the incidence and grade of glenoid radiolucencies. However, in a more recent study, an RM of the less than 4.5 mm was associated with an increased incidence of radiolucent lines and decreased patient-reported outcome scores at a mean follow-up of 41 months.¹¹

Optimal RM in ATSA remains an unanswered question. A complete understanding of optimal RM will require consideration of other factors including implant position and rotator cuff function.⁵

Early efforts to improve the stability of glenoid components and reduce the incidence of loosening led to the development of metal-backed implants. Most of these implants were uncemented and fixed to the glenoid with screws. These designs had a high failure rate, with early loosening, screw breakage, polyethylene dissociation, and the need for revision. Later efforts improved this design with the incorporation of a bony ingrowth material on the metal backing. Taunton et al examined the results of TSA with one such implant, reporting a 31% revision rate associated with loosening and an implant survival rate of 52% at 10 years. Porous ingrowth was improved and screw fixation was eliminated from the design. The metal backing was changed to a monoblock polyethylene platform with a central bony ingrowth attachment. However, these components fractured at the keel-glenoid junction or through the bone ingrowth platform, resulting in an unacceptably high failure rate. These implants were redesigned yet again, and the bony ingrowth platform and its connection to the polyethylene have been solidified to resist fracture. Modern metal-backed implants may hold promise; however, because of the history of loosening and catastrophic failure of early components, judicious use and close monitoring of these implants are recommended.⁵

All-polyethylene implants have proved to be a more durable option than metal-backed implants. This may be related in part to the favorable mechanical properties of all-polyethylene implants that result in decreased stress at the implant-bone interface. This is the reason why most glenoid implants used today are all polyethylene. Many of the advances in polyethylene processing and manufacturing associated with hip and knee arthroplasty have been applied to glenoid components. Cross-linked, ultrahigh-molecular-weight polyethylene has been shown to have favorable wear properties and a low incidence of wear-induced osteolysis. Backside texture of all-polyethylene cemented implants has also been examined, and rough texture and threading of pegs have been shown to improve implant pullout strength. Most all-polyethylene components are designed for cemented fixation around three or four pegs or a central keel. Several comparative studies have examined pegged and keeled implants and most have shown no significant difference in implant survival or clinical outcome despite a reported higher rate of radiolucent lines with keeled implants.⁵ In a biomechanical study, Roche et al reported the results of eccentric loading with similar RMs using a peg and keel glenoid and found no difference in edge displacement between the two designs.¹² In a prospective randomized trial comparing pegged and keeled components, Edwards et al showed that, even with the careful application of modern cementing technique, pegged components were radiographically superior to keeled glenoid components. At present the use of both keeled and pegged components can be supported based upon the clinical data reported.

In recent years, several variations to the traditional all-polyethylene pegged implants have been introduced all with the goal of improving implant stability. Divergent pegs have been used to provide additional stability against micromotion. All-polyethylene components that allow for bone ongrowth onto an interference-fit central peg provide the possibility of long-term biologic fixation. These “hybrid” components are characterized by central noncemented fixation and peripheral cemented fixation. Promising clinical results have been reported particularly when radiographic density is observed between the flutes of the central peg, indicating bone ingrowth. Other newer implants have been designed to take advantage of the bone ingrowth potential of porous-coated metal and the favorable mechanical properties of polyethylene by adding an optional porous-coated metal central post to a polyethylene-pegged component.⁵

Clinical and biomechanical studies have demonstrated excellent bone ingrowth but have also reported problems with metal debris formation, fracture, and/or dissociation at the metal-polyethylene interface. These complications have been accompanied by early revisions for aseptic loosening.¹³ Nelson et al. have reported a 36% rate of radiolucency lines at a minimum 5-year follow up on radiographic evaluation, when used exclusively hybrid glenoids after TSA. Nonetheless, they presented no cases with aseptic loosening.¹⁴ It has been also reported as little as less than 2% revision rates due to aseptic loosening when a hybrid glenoid is used.¹⁵ However, a recent matched cohort comparison, including same humerus component, reported a lower incidence of loosening compared to all-polyethylene glenoids (1.3% vs 3.8%) and lower revision rates with a minimum of 2 years of follow-up.¹³

In ATSA, the humeral component is responsible for a small number of complications and revision surgeries. In a study of 1423 patients who underwent ATSA from 1984 to 2004, Cil et al reported an 83% humeral component survival at 20 years.¹⁶ The early-design Neer II humeral component had a 98% survival rate. In a series of 1112 TSAs, isolated humeral loosening occurred in only 0.3% of cases. Given the relative rarity of isolated humeral loosening, if encountered, the surgeon must suspect and carefully evaluate for the possibility of a low-grade infection and potential other sources of the reported symptoms.⁷,¹⁶,¹⁷ The micromotion associated with humeral component loosening results in progressive endosteal bone erosion. The cortices become thinner; subsidence and progressive varus can result, ultimately leading to shaft fracture or perforation.