In patients with significant distal humerus bone loss , positioning of the humeral implant can be challenging, as there are few intraoperative anatomic landmarks from which to reference. In such cases, preoperative planning should include full-length radiographs of both the involved humerus and contralateral humerus. The contralateral humerus could be used to measure the total humeral length and a comparison is made to determine the amount of humeral bone of the affected humerus. Intraoperatively , humeral implant length should be based on soft tissue tension. This can be difficult in a flail elbow as soft tissue contractures are common. The “shuck test” can be used to determine humeral stem position when there is distal humerus bone loss. The humeral and ulnar components are placed and linked together. With the arm flexed to 90°, the forearm is distracted distally. This will result in optimal positioning of the humeral component [40]. If needed, the humerus can be shortened by up to 2 cm without causing triceps weakness [23].

Rotational alignment is also challenging with distal humerus bone loss. Inaccurate positioning can affect functional outcomes and lead to accelerated wear [41]. The proximal ulna can be used to assess for rotation alignment of the ulnar component by aligning the posterior aspect of the ulnar component with the flat dorsal aspect of the proximal ulna [42]. The humeral component should be internally rotated approximately 15° in relation to the posterior humeral shaft and, even in cases of severe bone loss of the distal humerus, the shaft can still be referenced [43]. Research has shown that image-based navigation leads to improved alignment compared to nonnavigated implantation although no currently available TEA system has the capability of image-based navigation [30].

Perforation of the canals is at high risk in cases of TEA after failed osteosynthesis. Cannulated flexible reamers can help avoid perforation. Placement of guidewires for reamers under fluoroscopy intraoperative can be used to confirm intramedullary placement and avoid perforation.

Wound closure for TEA should be meticulous as there is a relatively high rate of wound complications [27, 37]. If the triceps is reflected, it should be reattached through cruciate drill holes in the ulna [35]. The rate of triceps insufficiency after TEA is 2–3 % [36, 37]. The triceps should be completely repaired to the flexor and extensors isolating the joint from the subcutaneous region. This will help prevent hematoma or seroma development and limit the possibility for wound compromise. The subcutaneous tissue should be closed in layers and the skin should be reapproximated with interrupted nonabsorbable suture. A drain should be placed to prevent seroma or hematoma development in the subcutaneous tissue.

There is no consensus in the literature on postoperative management . Postoperative splinting in extension ranges from none to almost 2 weeks. Longer splinting time appears to be associated with decreased wound problems [37] but decreases postoperative range of motion. In the literature on TEA for elbow instability, splinting ranged from none to 2 days postoperatively with all splinting done in extension [19, 22, 23]. In general, anterior extension splinting should be performed to reduce tension and pressure on the posterior wound to prevent wound complications. For a triceps-on approach, immediate active and passive range of motion exercises are allowed, typically with nighttime anterior splinting to maintain extension. Patients are allowed immediate use of the limb for everyday activities of daily care after the initial splint is removed and then moved to the 5 lb lifting restriction at 6 weeks. If a triceps-off approach is utilized, active extension needs to be protected during the first 6 weeks after TEA to prevent triceps failure. Failure of the triceps to heal is a major complication associated with TEA. In cases of bone loss, exposure is not often a problem; therefore, a triceps-on approach is recommended if possible to prevent triceps failure.

Elbow instability is an indication for TEA for the carefully selected patient. Most reports of TEA for instability are for distal humerus nonunions in older patients. Early reports did not show favorable results with TEA [44, 45]. More recent studies using modern implants in older patients (over 60 years of age) demonstrate better outcomes (Table 16.1). The largest reported series is from the Mayo Clinic, which updated a previous series to include 91 consecutive patients with 92 elbows treated with TEA for distal humerus nonunion followed for an average of 6.5 years [18, 20]. The authors showed improvement in range of motion, elbow stability, and Mayo Elbow Performance Score (MEPS) in the majority of their patients. Based on the MEPS score, 78 % had a satisfactory (good or excellent) outcome postoperatively. They reported 44 complications, including infection, aseptic loosening, component fracture, periprosthetic fracture, and wound complications. Implant survival without removal or revision was 96 % at 2 years, 82 % at 5 years, and 65 % at 10 and 15 years [18].

Two smaller series out of Europe showed similar results to the Mayo Clinic series. Pogliacomi et al. followed 20 patients treated with TEA for distal humerus nonunion for an average of 65 months [22]. They reported improved MEPS score postoperatively with 90 % of patients having good or excellent outcomes despite a 30 % complication rate. Espiga et al. reported on six patients treated with linked TEA for symptomatic distal humerus nonunion followed for an average of 40 months [19]. They reported acceptable range of motion and improved pain postoperatively. Based on postoperative MEPS score, 67 % of the patients had a satisfactory outcome. There was one complication in their series, difficulty with wound healing that required fasciocutaneous flap coverage 2 months after surgery.

There is one series reported in the literature examining patients treated with TEA for distal humerus nonunion with a primary complaint of elbow instability [23]. Ramsey et al. reported on 19 elbows and showed improvement in MEPS score and range of motion postoperatively. There was no residual elbow instability postoperatively and 16 of the 19 elbows had satisfactory outcomes.

Complications from total elbow arthroplasty were summarized in two systematic review of primary TEA for all indications [36, 37]. The overall reported complication rate varied between 20 and 40 %. The rate of wound complications was 9 %, infection 4 %, and ulnar neuropathy 2–5 %. There was a 5–9 % loosening rate and 4 % rate of implant failure. Given the relatively high complication rate, TEA for instability should be reserved for older, functionally low-demand patients who have failed other treatment options.