Fractures of the Shoulder, Arm, and Forearm

Plate and screw fixation

Plating techniques continue to evolve. Newer precontoured plates allow more accurate fitting while maintaining strength; however, complications have been reported with 3.5-mm reconstruction plates, which allow easy contouring but may be too weak to maintain reduction. Currently, the most commonly used technique is superior placement of the plate ( Fig. 57.1 ) or anteroinferior plate placement because of the safe screw trajectory and less implant irritation ( Fig. 57.2 ). The recent trend of using smaller implants applies also to the clavicle. Evidence exists supporting the use of 2.7-mm compression type plating. The author reserves this technique when 2.7-mm plates are used in a neutralization technique with anatomic reduction and lag screw fixation. A report by Czajka et al. demonstrated a high rate of union with a low rate of soft-tissue irritation with the use of double mini-fragment plate fixation. This technique has become the most frequently used technique by the author ( Fig. 57.3 ). Regardless of the plate placement technique used, meticulous attention is mandatory to preserve the periosteum and avoid injury to the subclavian vessels and lungs; lag screw fixation should be used when possible.

Intramedullary fixation

Intramedullary nailing of clavicular fractures has been done for over 50 years, with a variety of devices, including Rockwood pins, Kirschner wires, Küntscher nails, and Rush nails ( Fig. 57.5 ). Suggested advantages of intramedullary fixation include small skin incision, less periosteal stripping, and relative stability to allow callus formation. Frequent complications, such as intrathoracic migration, pin breakage, and damage to underlying structures, however, have limited the use of this technique. A biomechanical study comparing fixation of clavicular osteotomies with 3.5-mm compression plates and 3.8- or 4.5-mm intramedullary pins also showed that plated constructs were superior in resisting displacement. More recently, titanium elastic intramedullary nails have been used, with good results reported in a number of studies. However, reported complication rates have ranged from 9% to 78% with these devices, mainly medial or lateral migration and perforations. Frigg et al. reported a reduction in complications from 60% to 17% with the use of an end cap, converting to open reduction after two failed attempts at closed reduction, using careful manual passage of the nail, obtaining intraoperative oblique radiographs to rule out lateral perforation, and limiting postoperative range of motion to 90 degrees for 6 weeks.

Treatment options

Almost all scapular body and neck fractures are still treated nonoperatively. We immobilize the shoulder for 2 to 3 weeks and begin an active-assisted range-of-motion protocol when pain permits. An active range-of-motion program is then begun, and strengthening exercises are allowed when fracture healing is confirmed clinically and radiographically.

The mobility of the shoulder is predictive of function in many patients with a scapular fracture; however, there is still a small group of patients in whom ORIF probably is indicated. The goal of treatment is to preserve shoulder function by avoiding malalignment, arthrosis, scapulothoracic dyskinesis, and impingement pain ( Table 57.7 ).

Glenoid fractures

Glenoid fractures should be treated as all other intraarticular fractures, and reduced and stabilized when significant (>4 mm) displacement exists through the articular surface that leads to joint subluxation or incongruency. Anavian et al. reported that, of 33 patients with complex and displaced intraarticular glenoid fractures, 87% were pain free and 90% returned to preinjury levels of work or activity after operative treatment. The operative approach of choice is the Judet or modified Judet (see Technique 1.94). Additional anterior approaches occasionally are needed.

Scapular body or neck fractures

Fractures of the scapular body or neck that are so significantly displaced that malunion and pain are of concern should be considered for operative treatment. Medialization of the glenoid has been questioned by Zuckerman et al., who recommended evaluation of lateralization of the scapular border. CT evaluation also found that in patients with glenoid neck fractures, pure medial translation of the glenoid relative to the axial skeleton was rare; instead, there was typically a component of shortening of the scapular width combined with lateralization of the scapular body. Treatment decisions should be based on the amount of displacement. Some authors use the glenopolar angle as a criterion for determining treatment. This angle is formed by a line drawn from the inferior pole of the glenoid fossa up to the superior pole and a second line drawn from the superior pole of the glenoid fossa down through the inferiormost angle of the scapular body ( Fig. 57.14 ). The normal glenopolar angle ranges from 30 to 45 degrees. Anavian et al. suggested that three-dimensional CT is more reliable than plain radiography in the evaluation of extraarticular scapular fracture displacement.

Cole et al. listed several criteria for operative treatment of scapular fractures:

• ▪

  A 15 to 20 mm lateral border offset (lateralization)

• ▪

  Forty degrees of scapular body angulation, as measured on a scapular-Y view

• ▪

  Glenopolar angle of 20 degrees or less and greater than 60 degrees.

• ▪

  Scapular body fracture with injury to the clavicle or clavicle-acromion complex.

These decision-making criteria have not yet been shown to produce improved outcomes; the surgeon’s skill level and patient issues should contribute to the decision for operative treatment. We generally favor conservative treatment, but this is an active treatment decision and not “benign neglect” ( Fig. 57.15 ).

Classification

The most commonly used classification system for proximal humeral fractures is that of Neer ( Fig. 57.16 ). Although limited reliability, reproducibility among observers, and consistency by the same observer at different times have been cited as limitations of the Neer system, it remains useful in guiding treatment. Classification is based on the four-part anatomy of the proximal humerus: the humeral head, the lesser and greater tuberosities, and the proximal humeral shaft. The criterion for displacement is more than 1 cm of separation of a part or angulation of 45 degrees. Displaced three-part and four-part fractures markedly alter the articular congruity of the glenohumeral joint and have the highest likelihood of disrupting the major blood supply to the proximal humerus ( Fig. 57.17 ). Osteonecrosis is most likely after displaced four-part fractures.

Radiographic evaluation

An anteroposterior view of the shoulder in the plane of the scapula, a lateral view of the scapula (Y view) ( Fig. 57.18 ), and a supine axillary view ( Fig. 57.19 ) are necessary in all patients initially to evaluate a proximal humeral fracture. If the amount of displacement of the humeral head or tuberosity fragments is unclear on radiographs, an axial CT scan with 2-mm sections is indicated ( Fig. 57.20 ).

Nonoperative treatment

Nonoperative treatment can obtain a functional, painless extremity in most proximal humeral fractures. The range of motion of the shoulder joint accommodates moderate angular deformity without significant functional loss. Neer described acceptable angulation as less than 45 degrees and less than 1 cm of displacement. Although these criteria are not absolute, they do provide a guide. An elderly, infirm patient can tolerate functional loss better than a young, active patient. The first step in treatment decision-making is to determine if displacement (<66%) and angulation (varus is poorly tolerated) are acceptable for a particular patient; the second is to determine if the humeral head and shaft move as a unit. If both of these conditions are present, the fracture is stable and in an acceptable position. A sling is used for comfort, and a physical therapy regimen with pendulum exercises is started, usually within 1 week. If the humeral head and shaft do not move as a unit, physical therapy can be delayed for 2 to 4 weeks in patients who are poor surgical candidates because of age, low functional demands, or comorbidities that preclude participation in rehabilitation. In young, active patients, early operative fixation should be considered. Generally, the longer the period of immobilization, the longer the period of therapy, and the greater the disability. A randomized controlled trial involving 74 patients with impacted proximal humeral fractures found that early (within 72 hours of injury) passive mobilization was safe and more effective in restoring function than conventional immobilization (3 weeks) followed by physical therapy. Another study, however, pointed out that fracture settling continues to occur with conservative treatment.

Operative treatment

The decision that operative treatment is appropriate is complicated by the numerous and varied techniques described for fixation of proximal humeral fractures. Generally, fracture displacement is used as the indicator of stability. The goal is restoration of proximal humeral anatomy with stable fixation that allows early functional range of motion. Chronic malunions and nonunions that are subsequently treated surgically are associated with poor outcomes. Consequently, it is imperative to recreate the normal proximal humeral anatomy with respect to tuberosity reduction and the head-neck relationship. Indications for operative treatment include displaced two-part surgical neck fractures, displaced (>5 mm) greater tuberosity fractures, displaced three-part fractures, and displaced four-part fractures in young patients. The type of fixation (transosseous suture fixation, percutaneous pinning, intramedullary nailing, or plate fixation) used depends on the patient’s age, activity level, and bone quality; the fracture type and associated fractures; and the surgeon’s technical ability ( Table 57.8 ). Age alone has been shown both to be predictive of failure and to have no association with failure. In their series of 154 fractures with proximal humeral fractures treated with plating, Boesmueller et al. found that the risk of screw cut-out was four times higher in patients over the age of 60 years and the overall risk for complications was three times higher than in younger patients.

Before surgery is considered, it is important to determine if the blood supply and bone quality are adequate. The Hertel radiographic criteria for perfusion of the humeral head ( Fig. 57.21 ) can be used to predict ischemia: metaphyseal extension of the humeral head of less than 8 mm and medial hinge disruption of more than 2 mm are predictive of ischemia. The combination of metaphyseal extension of the humeral head, medial hinge disruption of more than 2 mm, and an anatomic neck fracture pattern has a 97% positive predictive value for humeral head ischemia. According to the AO/ASIF classification system, extraarticular type A fractures have an intact vascular supply, whereas type B fractures have a possible injury to the vascular supply and type C articular fractures have a high probability of osteonecrosis. The cortical thickness of the humeral diaphysis has been suggested to be a reliable and reproducible predictor of bone mineral density and the success of internal fixation. The combined cortical thickness is the average of the medial and lateral cortical thickness at two levels ( Fig. 57.22 ). Generally, a cortical thickness of less than 4 mm precludes internal fixation because adequate screw purchase cannot be obtained; sling immobilization, transosseous suture, or hemiarthroplasty may be better options. Spross et al. and Newton et al. also demonstrated that the quality of the bone was associated with late cut-out.

Transosseous suture fixation techniques are well defined in the orthopaedic literature. Park et al. reported 78% excellent results in patients with two-part and three-part proximal humeral fractures treated with suture fixation. The use of strong nonabsorbable suture provides the advantage of incorporating the rotator cuff insertion to increase fixation in patients with poor bone quality ( Fig. 57.23 ). The level of soft-tissue dissection is not extensive, and relatively low rates of osteonecrosis have been reported with this technique. Concerns include the ability of the patient to move the shoulder joint and loss of reduction secondary to a nonrigid construct. More recently, Dimakopoulos et al. reported good results in 188 displaced proximal humeral fractures treated with transosseous fixation ( Fig. 57.24 ). They suggested as advantages of this technique less surgical soft-tissue dissection, a low rate of humeral head osteonecrosis, fixation sufficient to allow early passive joint motion, and the avoidance of bulky and expensive implants.

Percutaneous pinning has the advantage of avoiding further damage to the soft-tissue envelope and the blood supply to the humeral head ( Figs. 57.25 and 57.26 ). It also is a relatively inexpensive technique, and several series have reported good results in two-part, three-part, and valgus-impacted four-part fractures. The procedure is technically challenging and requires a satisfactory closed reduction, adequate bone stock, minimal comminution (particularly of the tuberosities), an intact medial calcar, and a compliant patient. In their series of 74 older patients (average age, 71 years), Calvo et al. demonstrated that reduction was associated with satisfactory outcome. However, if satisfactory closed reduction cannot be obtained, another form of reduction and fixation should be used. Loss of fixation, pin track infections, and axillary nerve injuries are common complications. Terminally threaded Schanz pins and bicortical pins inserted from the greater tuberosity to the medial humeral shaft add stability to the overall construct. Percutaneous pinning is contraindicated for fractures with metaphyseal comminution.

Intramedullary nailing (SEE Technique 57.4) provides more stable fixation than percutaneous pinning, although less than locked plate fixation. The Polarus nail (Accumed, Portland, OR) has been shown to provide more biomechanical stability than pin fixation, and good clinical outcomes have been reported with this device. Newer nail designs with polyaxial screws have more stability than earlier designs, and the addition of polyethylene bushings may increase stability and prevent screw back-out ( Fig. 57.27 ). Insertion of an intramedullary nail into the proximal humerus violates the rotator cuff, which can lead to postoperative shoulder pain. The advantages of the technique include preservation of the soft tissues and the theoretical biomechanical properties of intramedullary nails. A comminuted lateral cortex fracture or fractures involving the tuberosities may be a contraindication to intramedullary nailing. A recent randomized controlled trial demonstrated that complications were fewer with a straight nail design compared with a curvilinear design. A systematic review by Wong et al. reported satisfactory results in displaced two- and three-part proximal humeral fracture treatment with intramedullary nails. Sun et al. compared locking plates with intramedullary nails in displaced proximal humeral fractures in a systematic review and meta-analysis and demonstrated similar performance between the two fixation types.

Plate-and-screw constructs provide the most stable fixation of the three fixation methods ( Fig. 57.28 ). Locked plates add stability, especially in osteoporotic bone. An open reduction and rigid fixation allow accurate reduction and stabilization of the tuberosities, which is important because malunion of the tuberosities is poorly tolerated and is associated with poor outcomes in posttraumatic reconstructive shoulder arthroplasty. A prospective randomized trial by Zhu et al. found that at 1-year follow-up patients treated with locking plates had better outcomes than those treated with locked intramedullary nailing, but at 3-year follow-up outcomes were equal. The locking nail group had a significantly lower complication rate (4%) than the locking plate group (13%). Konrad et al. also reported similar outcomes in three-part proximal humeral fractures treated with intramedullary nailing (58 fractures) or plate fixation (153 fractures).

Historically, plate fixation of the proximal humerus has been fraught with complications, with malunion and nonunion caused by poor fixation in the humeral head ( Fig. 57.29 ). In addition, extensive soft-tissue dissection increases the possibility of osteonecrosis of the humeral head, leading to a painful and functionally limited shoulder joint. The development of locked proximal humeral plates was expected to improve treatment of these complex injuries greatly. The advantage of ORIF with a locked plate is an ability to reduce the fracture fragments into an anatomic position and stabilize them rigidly to allow early motion. Numerous outcome studies are now available because the locked proximal humeral plate has been widely used for more than 10 years; however, as was pointed out in a Cochrane review, there is little level I or II evidence. A recent randomized controlled trial comparing locked plating with conservative treatment of three-part and four-part fractures in elderly patients found no difference in outcomes at 1-year follow-up. Despite the lack of a large body of supporting literature, the locked proximal humeral plate is considered by most fracture surgeons to be a great improvement in the management of proximal humeral fractures, and it has become the implant of choice for these fractures. Schnetzke et al. reported 98 patients treated with locked plating and concluded that anatomic reduction significantly improved outcomes.

Much attention has been focused on the medial side of the metaphyseal injury. Gardner et al. called attention to this by documenting the importance of the inferior screw behaving as a medial calcar substitution. Biomechanical studies have confirmed this importance, and Jung et al. confirmed it clinically by identifying medial comminution and insufficient medial support (no cortical or screw support) as independent risk factors for loss of reduction in 17 (7%) of 252 proximal humeral fractures. As an alternative to medial calcar screws, fracture site impaction adds stability by impacting the humeral head onto the humeral shaft. As modified by Torchia, valgus impaction osteotomy ( Fig. 57.30A-D ) appears promising, although no large series have been reported. In a biomechanical study, Weeks et al. found that fracture impaction increased the ability of the locking plate to withstand repetitive varus loading and was biomechanically superior to locking plate fixation alone. Gardner et al. described the use of a fibular strut graft to provide medial column support. Although promising, the technique also is demanding, and further randomized trials are needed to confirm its efficacy. Kim et al. noted improvement using a fibular strut graft versus inferomedial screws in conjunction with locking plates in four-part fractures, but there was no advantage noted in three-part fractures. In a systematic review, Saltzman et al. found satisfactory results when fibular strut grafts were used for augmentation.

Some issues with open reduction and locked plating include the extensive exposure required for plate application that carries a risk of damage to neurovascular structures, especially the ascending branch of the lateral circumflex artery. The complication and reoperation rates do remain high with this technique. Screw perforation through the humeral head is the most frequently reported complication. Perforation can occur as cutout from fracture settlement or from poor initial technique. Calcium phosphate cement augmentation has been shown to decrease this complication. Other complications include arthrofibrosis, impingement, malunion, nonunion, osteonecrosis, infection, and hardware failure. Poor outcomes are associated with initial varus displacement of three- and four-part fractures.

In an attempt to decrease complications with plate fixation, Gardner et al. used an anterolateral acromial (Mackenzie) approach in which the axillary nerve is identified and protected, anterior dissection near the critical blood supply is avoided, substantial muscle retraction is minimized, and the lateral plating zone is directly accessed (SEE Technique 57.6). Laflamme et al. reported no axillary nerve injuries and no loss of reduction in fractures treated with percutaneous humeral plating through two minimal incisions (a lateral deltoid split and a more distal shaft incision). As our understanding of the anatomy of the proximal humerus and our instruments improve, less invasive techniques appear promising. Electrophysiologic findings in a study by Westphal et al., however, revealed a 10% axillary nerve injury rate.

Complications

The most common complication of proximal humeral fractures is loss of motion (stiffness). Early physical therapy is associated with improved motion, but many patients do not recover full motion even with early physical therapy. Impingement from high-riding tuberosities or subacromial scarring also can limit motion. Nonunion also is fairly common, but nonunion rates have been decreasing with the use of new technologies such as locking plates and improved intramedullary nails. Malunion can result from unstable or delayed fracture fixation, patient factors, and poor surgical technique. In older patients with limited functional demands, malunion generally is well tolerated, but it may be debilitating in younger patients because of poor shoulder function, impingement, or rotator cuff tears. Osteonecrosis is relatively uncommon after nondisplaced or unoperated two-part and three-part fractures; functional outcome is improved if the proximal humeral anatomy has been restored. The presence of osteonecrosis does not always result in a poor outcome; osteonecrosis may be evident radiographically but cause minimal symptoms. Because late hemiarthroplasty has poorer results than early hemiarthroplasty, it is important to be sure that ORIF can adequately stabilize four-part fractures and restore humeral anatomy before this option is chosen.

Implant choice

The most commonly used plate for fixation of humeral shaft fractures is the broad, 4.5-mm, limited-contact dynamic compression plate ( Fig. 57.41 ); occasionally, a narrow, 4.5- or 3.5-mm, limited-contact dynamic compression plate is used for smaller bones. The distal metaphyseal-diaphyseal transition zone may require dual 3.5-mm, limited-contact dynamic compression plates ( Fig. 57.42 ) or newer plates designed specifically for the metaphysis. For spiral or oblique fractures, the ideal construct consists of a lag screw with a neutralization plate, whereas transverse fractures are ideally suited for a compression plating technique. In these fractures, attaining provisional reduction with a lag screw, Kirschner wire, or mini-fragment plate (Eglseder technique) allows direct observation of the reduction and a relatively simple plate application on the reduced humeral shaft ( Fig. 57.43 ); we believe this also limits periosteal stripping by clamps.

Comminuted fractures may require a bridge plating technique. Anatomic reduction of each fracture fragment is unnecessary. Attaining correct alignment, rotation, and length without disrupting the soft-tissue attachments to the comminuted fragments often leads to successful healing. Livani et al. reported 15 patients with bridge plating done through two small incisions proximal and distal to the fracture; all fractures united within 12 weeks except for a grade III open fracture with an associated brachial plexus injury.

In patients with poor bone quality, longer implants should be used to improve stability ( Fig. 57.44 ). Locking plates and screw augmentation with methyl methacrylate have been reported to add more stability to the construct. Generally, at least eight cortices (four screws) above and below the fracture are necessary to avoid screw pullout. The length of the plate is as important as the number of screws. More screws and longer plates for a greater working length of the implant may be needed for instability caused by poor bone quality or fracture comminution. We reserve the use of locking screws for poor bone quality and short segments.

Approach

Numerous approaches can be used for plate fixation of the humerus. Fractures of the middle or proximal third usually are best approached through an anterolateral approach (brachialis-splitting approach). A posterior approach (triceps-splitting or modified posterior approach) is best for fractures that are midshaft or extend into the distal third of the humerus ( Fig. 57.45 ). Gerwin, Hotchkiss, and Weiland described a modified posterior approach in which the triceps is reflected medially off the lateral intermuscular septum (SEE Technique 57.7). This approach exposes an average 10 cm more of the humeral shaft than the standard posterior approach. Less frequently, a direct lateral or anteromedial approach may be appropriate. A recent study using this approach showed high union rates and low complications.