20 Component Loosening: What Can Be Done to Prevent and Treat It
This chapter discusses the incidence of loosening of humeral and glenoid components along with considerations for optimizing fixation in both primary and revision surgery. The successful use of both cemented and uncemented humeral components has been reported in the literature. Modern baseplates and glenospheres have a low incidence of loosening, but screw configuration and bony support play key roles in longevity. There are unique considerations with respect to both the humerus and the glenoid when bone loss is present. Advances in implant design as well as modular components may optimize implant fixation where bone graft and cement have had poor long-term outcomes.
Historically, concerns for loosening in shoulder arthroplasty have centered on the glenoid component. Current literature has provided significant variances in the rate of glenoid loosening, with rates from as low as 0 to 12%, up to 96% when considering radiolucent lines as a sign of early loosening. 1 Humeral components in primary, unconstrained, anatomic shoulder arthroplasty have very low rates of loosening or radiolucency. Press fit stems, with and without porous coating or other bone integration technologies have been well reported to have low incidence of loosening and radiolucency. 2 , 3 While early revision rates remain low across all types of shoulder arthroplasty, the most common complications necessitating revision include dislocation (twice as likely with reverse versus anatomic arthroplasty), loosening (twice as likely with anatomic versus reverse arthroplasty), and infection (with similar rates in anatomic, hemiarthroplasty, and reverse arthroplasty). 4
With consideration of the reverse shoulder arthroplasty, initial attempts at constrained total shoulders for soft-tissue deficiency and instability generally failed early given that forces across the joint due to constraint and deltoid tension proved too great for many early devices. 5 The Grammont design allowed for the success of a constrained device and modern reverse shoulder arthroplasty was born. By medializing the center of rotation, forces on the glenoid implants were neutralized and early catastrophic failure avoided. The addition of locking screws to the baseplates and various bony ingrowth or ongrowth technologies have afforded continued success of the reverse shoulder arthroplasty for many complex reconstruction situations.
However, there continue to be concerns for loosening and component failure. Excluding notching as a cause, both the glenoid component and the humeral component can be subject to loss of fixation, loosening, and failure.
20.2 Humeral Stems
Early reverse total shoulder designs relied on cemented humeral stems, including both monoblock and modular designs. With the use of modern reverse total shoulder arthroplasty, the rate of humeral loosening is generally low. A recent publication examining uncemented trabecular metal porous coated implants for reverse total shoulder arthroplasty demonstrated 10% rate of lucent lines, with none involving more than one zone or having more than 2 mm of lucency. 6 They also reported a low rate of subsidence, with only 9% showing subsidence of < 2 mm at 1 year with no progression at 2 years. 6 When evaluating the rates of humeral stem radiolucency involving three to four zones, Melis et al found significantly higher rates in cemented stems (20%) versus uncemented stems (8%). 7 However, it was also noted that radiographic signs of stress shielding were more frequent in uncemented components, as was resorption of the greater and lesser tuberosities. 7 Comparing cementless versus cemented fixation with a grit blasted stem, a loosening rate of 2% (cementless) and 3% (cemented) was observed. 8 These findings were corroborated by Wiater et al as they found no evidence of loosening or humeral components at risk of loosening across 64 cementless and 37 cemented Grammont-style reverse total shoulder arthroplasties. 9 Several other studies have shown little to no radiographic evidence of stem loosening in uncemented stems, including mini or stemless devices. 9 , 10 , 11 , 12 , 13 In addition, nontraditional short stem or stemless devices have published series with no radiographic or clinical loosening. 14 Ultimately, there is support from current literature that uncemented humeral stems carry a low rate of loosening.
There are significant compressive and rotational forces on the humeral component. Rare complications have been reported in the past with modular components including unscrewing of the metaphysis from the diaphysis. 15 In our practice, uncemented humeral components have been utilized as frequently as possible, with excellent results and few complications. Intraoperatively, we use a “twist test” on the broach, confirming a solid fit, to ensure the size is appropriate. Our implant of choice is a humeral stem with a proximal metaphyseal porous coating for additional fit and eventual bony ingrowth. We exercise caution with the use of short-stemmed devices (< 75 mm). While we agree that this may be a sound option for the younger patient with appropriate pathology who may require a revision, we have had one case of medial subsidence after placing a short-stemmed device in a patient who was felt to have adequate cancellous bone density. Once the medial portion of the stem rested completely on the calcar, subsidence stopped, and no revision has been required at 2 years (► Fig. 20.1).
Occasionally, due to either poor bone quality or revision implantation, cement is still required. The use of cement fixation in arthroplasty has been extensively studied in the adult reconstruction literature; however, many of the findings and recommendations are applicable to shoulder arthroplasty. As with arthroplasty in any joint, the utilization of proper cementing technique is of paramount importance to ensure adequate fixation and prevention of failure at the bone–cement and/or the stem–cement interface. Cement fixation also carries unique medical complications. Recent literature has described significant drops in mean arterial pressures, systolic and diastolic blood pressures, pH and base excess decreases, and occasional arrhythmias during cementation. 16 Ultimately, cementation of shoulder arthroplasty can be safe and effective given adequate anesthesia observation intraoperatively. We use a polished humeral implant with an asymmetric body for rotational stability. We cement using standard modern technique with a bioabsorbable canal stop and pressurized medium-viscosity cement. We have had two cases of cemented stems exhibit early (5 years or less) loosening and both were revisions of initially loose components. Both cases had proximal humeral bone loss resulting in a cylinder being cemented into a tube (► Fig. 20.2). In general, we find that the degree of proximal humeral bone loss contributes most significantly to humeral component loosening. While this generally is the worst-case scenario, it may be successfully managed with long-stemmed components as demonstrated by Stephens et al. 17 More recently in the setting of revision fixation, we have transitioned to a modular system that can be press fit in the diaphysis and built to the appropriate length for the patient. Results for this technique are short term, but securing the stem by solid press fit and ingrowth in the distal diaphyseal segment has advantages. Longer-term study is needed to make a definitive recommendation on cemented versus uncemented revision components and will likely depend on multiple factors such as degree of bone loss and length of the implant.