30.2.1 Infection

As with any failed arthroplasty, there has to be a high level of suspicion of an infection. It has been shown that low-grade infection can be difficult to diagnose, particularly with the prevalence of Propionibacterium ⁴ in failed shoulder arthroplasty. Studies have shown that infection has been the cause of some failed TWA but it does not seem to be as much common as in the shoulder. ⁵ The highest concentration of P. acnes has been demonstrated around the shoulder girdle and axilla and not over the wrist; however, it remains a potential source. ^{6 , 7} Perioperative tests for infection in revision arthroplasty can be challenging with high sensitivity but low specificity.⁴ Common practice to exclude/diagnose infection is the combination of clinical suspicion, blood markers, and in some situations an aspiration or tissue biopsy for microbiological culture. If there is a suspicion of infection and loosening, then similar to other arthroplasties a revision can be performed as a two-stage process rather than a simple reimplantation. The two-stage process involves the removal of the suspected infected implant with the placement of a temporary antibiotic-loaded spacer. Following a period of time typically 6 to 8 weeks depending upon the organism present and antibiotic sensitivity, a second stage is performed with a reimplantation of the prosthesis.

30.2.2 Aseptic Loosening

Implant loosening in the absence of infection can be the result of inadequate initial fixation, mechanical loss of fixation over time, or secondary to a particle-induced osteolysis resulting in biologic loss of fixation. Loosening or bone reabsorption can be due to the forces exerted by the implant on the surrounding bone. Loosening of a TWA can cause pain and instability which may require a revision arthroplasty (Fig. 30.1, Fig. 30.2).

At our institution we reviewed 220 TWA, predominantly patients with rheumatoid arthritis (92%). Twenty-five wrists were identified as requiring revision surgery. The most common cause of revision after a TWA is implant loosening of the distal component in our series. ¹¹ Work done by Retting and Beckenbaugh showed a comparable pattern of loosening, with the majority needing revision for loosening of the distal component. ¹²

Boeckstyns and Herzberg looked at 44 consecutive cases of Re-Motion TWA (Small Bone Innovations Inc.). They found periprosthetic radiolucency (more than 2 mm) around the radial component in 16 cases and on the carpal side in 7 cases (Fig. 30.3, Fig. 30.4). These developed gradually, adjacent to the prosthetic components regardless of the primary diagnosis, and seemed to stabilize in most patients over a period of 1 to 3 years. They found that in a small percentage of patients the periprosthetic area of bone was significantly larger. In their study they were able to show that the radiolucency was not necessarily related to loosening of the implants; only five carpal components and one radial component subsided or tilted. They recommended close and continued observation of TWA and especially radiolucency. A similar phenomenon has been demonstrated in ulnar head replacements in which no artificial components articulate with each other. This resorption seems to stabilize after a period of 6 months to 1 year and is attributed to stress shielding. ¹³

Packer et al have published promising intermediate results using the Motec TWA which is a cementless modular metal-on-metal ball-and-socket wrist arthroplasty. Rokkum et al using the same arthroplasty reviewed 57 patients with end-stage arthritis. They report survivorship of 86% at 10 years, suggesting that the Motec TWA can provide long-lasting unrestricted hand function in young active patients. Recently, Reito et al reported early catastrophic failure in two cases. In one case, the failure was secondary to metal debris (Fig. 30.5) and in the other it was thought to be due to an adverse reaction to polyether ether ketone. The first patient showed elevated levels of blood cobalt and chrome levels and magnetic resonance imaging (MRI) imaging showed clear signs of a pseudotumor. The other patient had an extensive release of polyether ether ketone particles into the surrounding synovia due to adverse wear conditions in the cup, leading to the formation of a fluid-filled cystic sac with a black lining and diffuse lymphocytic dominated inflammation in the synovia. The authors recommended regular follow-up including radiographs and monitoring of cobalt and chrome ion levels and a low threshold for cross-sectional imaging in patients who underwent TWA with a Motec TWA.

The biaxial and Universal II implant has five zones described in the distal radial area and six zones in the metacarpal and carpal areas (Fig. 30.6).

30.2.3 Implant Fracture

This is an uncommon presentation secondary to implant loosening. In our series we had three implant fractures: two with breakage of the screws used to fix the distal component, and one component a fracture of the peg at its junction with the baseplate (Fig. 30.7). All of the prostheses could be salvaged with revision TWA. ²

30.2.4 Biomechanical Mismatch

Like all unconstrained prostheses, the success of a TWA is dependent on the remaining bone stock. Conservation of bone stock is particularly important in patients with rheumatoid arthritis and adequate soft tissue releases are crucial to ensuring an adequate exposure and to help balance the prosthesis. In our experience, volar capsular releases and tendon transfers are typically required in the presence of a severely deformed wrist. This is based on our knowledge in dealing with severe deformities in primary wrist replacements, although our experience with revision surgery is limited in this respect. In our series of patients, one patient underwent revision at 4 months postoperatively due to inadequate distal radial resection. ¹⁵ Vogelin and Nagy reviewed a series of failed Meuli TWA. They reported the failures occurred due to a combination of mechanical failure and soft tissue problems. ¹⁴ Stringer et al in his series of 23 Swanson silicone rubber implants used in patients with rheumatoid arthritis published an implant fracture rate greater that 50% with a revision rate of 30%. Particulate tenosynovitis and instability with radiologic deterioration were reasons for revision surgery. The Swanson wrist arthroplasty is largely of historical significance.

30.3.1 Bone Grafting

At revision TWA there may be a considerable bone loss particularly around an otherwise well-fixed implant. A large bone defect makes replanting a prosthesis challenging. Gaining stability of an implant in the absence of a solid bone fixation is likely to lead to early failure. The options available for filling of the defect include autograft and allograft bone augmentation. Autograft bone stock has limited supply whilst allograft bone stock has a much larger volume available but has a low risk of infection and is biologically inert. Bone grafting of the defect can be carried out in a single- or two-stage process. Stability can be a challenge with a single-stage bone grafting in large defects. In a two-stage procedure, bone is impacted into the defect and allowed to incorporate, and then the reimplantation is performed at a later stage typically 8 to 12 weeks when the bone is thought to have incorporated. Incorporation is assessed radiologically when the graft appears to merge with the adjacent cancellous bone.

30.3.2 Wrist Arthrodesis

Wrist fusion is a recognized and recently more popular salvage procedure for failed TWA. ¹⁷ A fused wrist provides a stable wrist joint but it can be challenging due to the difficulties in restoring wrist height, obtaining a stable fixation, and achieving bony fusion. The fusion procedure is routinely augmented with bone graft to promote healing. The remaining articular cartilage is prepared to allow bony union. During this osteointegration phase, stability is commonly maintained with a dorsal plate running between the base of a metacarpal and the distal radius. Carlson and Simmons ¹⁷ showed that in their series of 12 failed wrist arthroplasties they were able to achieve fusion in all patients. They combined a bulk allograft from a femoral head and iliac crest autograft and were able to achieve a largely pain-free fusion. ¹⁴ Although these results are promising, the sacrifice is the loss of wrist movement (Fig. 30.8). Interposition bone graft is a useful technique to regain wrist length to optimize the function of the flexors and extensor tendons, thereby restoring grip strength. However, in the presence of extensive bone loss overlengthening can cause issues with the median nerve and loss of tendon excursion due to a fixed length phenomenon, not dissimilar to what occurs in a chronic compartment syndrome.