Practice Essentials
The shoulder links the upper extremity to the thorax. Optimal functioning of the upper extremity requires mobility and power that allow a range of performance, from powerful, explosive movements (eg, throwing a baseball 100 mph) to very accurate, fine movements (eg, performing microsurgery or playing the violin). Tasks of daily independence require the ability to position the hand throughout the range of an imaginary sphere.
In addition to limiting function, disorders of the shoulder can cause pain, which, in turn, can affect the patient's work and sleep. Therefore, fractures of the proximal humerus can be devastating to quality of life. These fractures can also cost society a significant loss of productivity from otherwise viable members of the workforce.
Hippocrates first documented a proximal humerus fracture in 460 BCE and treated it with traction. In 1869, to improve treatment, Krocher classified fractures of the proximal humerus. In 1934, Codman developed a classification that divided the proximal humerus into four parts on the basis of epiphyseal lines. In 1970, Neer's classification expanded on the four-part concept and included anatomic, biomechanical, and treatment principles, providing clinicians with a useful framework for diagnosing and treating patients with these fractures. [1]
Successful treatment of fractures of the proximal humerus (ie, that portion involving the glenohumeral articulation) presents a challenge for physicians. Many factors must be considered in developing a treatment plan. Accurate assessment of the fracture, patient compliance, medical comorbidities, and time from injury to treatment are critical factors affecting outcome. Additionally, technical factors in the reconstruction of these fractures require surgical experience that few surgeons have the opportunity to develop.
Indications for treatment are displaced articular fractures and periarticular fractures. However, the "personality" of the fracture (eg, bone quality, fracture orientation, concomitant soft-tissue injuries), the personality of the patient (eg, compliance, realistic attitude, mental status), and the personality of the surgeon (eg, surgical experience, technical familiarity, available resources) all have a tremendous effect on specific treatment indications.
Contraindications for repair of proximal humerus fractures include inability to tolerate the procedure medically and lack of clearance for surgery through the primary care physician or specialty consultants.
Initially, treatment of proximal humerus fractures consisted of closed reduction, traction, casting, and abduction splints. In the early 1930s, operative treatment for displaced fractures gained popularity, which continued in the 1940s and 1950s. Humeral head replacement for severely displaced fractures of the proximal humerus was introduced the 1950s. In the 1970s, the AO/ASIF (Arbeitsgemeinschaft für Osteosynthesefragen/Association for the Study of Internal Fixation) group popularized plates and screws for fracture fixation, and humeral head prostheses were redesigned.
Currently, fixation methods that involve limited fixation and limited dissection are becoming more popular, and prosthetic replacement for severe fracture is being refined further. [2, 3]
Anatomy
Osteology
The anatomy of the proximal humerus is highly variable. Multiple cadaveric studies have been performed to compare anatomic relations that are constant among individuals. Unfortunately, few exist. The critical anatomic relations of the proximal humerus are those of the articular segment to the shaft and the tuberosities. These include retroversion, inclination angle, and translation of the head relative to the shaft, as well as the relation of the head to the greater tuberosity.
On average, the articular segment is retroverted 30° relative to the forearm. The range is quite large (0-70°) and can vary from one side to the other. Inclination of the articular segment also can vary (from 120º to 140°).
The head segment can lie directly over the medullary canal but often is translated either posteriorly or medially. Therefore, if a prosthetic replacement is placed in the intramedullary canal, a resultant shift in position of the articular segment can occur unless some design feature of the prosthesis allows for a simultaneous shift in the prosthetic head position. Finally, the proper anatomic relations of the prosthetic head must be reconstructed meticulously to avoid overreducing the tuberosity to the head height. [4]
The articular head always lies above the greater tuberosity, but the difference can range from 3 to 20 mm. The biceps groove at the level of the articular surface has a constant relation to the version of a prosthetic articular surface in relation to the fins of the prosthetic body. If the anterior fin is placed at the biceps groove, the articular segment will be in 30° of retroversion. If the posterior fin is placed 8 mm posterior to the biceps groove, the same degree of retroversion will be recreated.
Injury to the blood supply of the proximal humerus has been implicated in the development of avascular necrosis (AVN). [5] The ascending branch of the anterior circumflex humeral artery (artery of Liang) was found by Gerber to provide most of the blood flow to the articular segment. If the medial calcar of the humerus is spared by the fracture, the vessel will be spared. Hettrich et al, however, have suggested that the posterior branch is more important. [6]
Rotator cuff
The rotator cuff is the critical structure that must be reconstructed following proximal humerus fracture. The initial fracture pattern, displacement of the fracture fragments, reduction maneuvers, and fixation techniques used to oppose the displacement forces are dependent on the rotator cuff forces that produced the fracture. [7]
The supraspinatus attaches to the greater tuberosity at the superior facet and the superior half of the middle facet. Avulsion-type forces from this muscle produce a short transverse fracture of the greater tuberosity that displaces primarily superiorly. Straight abduction helps reduce the fragment, and tension band fixation neutralizes initial displacement forces.
If the infraspinatus, which attaches to the entire middle facet of the greater tuberosity, also is involved, the fracture fragment is larger, and the fragment is displaced posterosuperiorly. In addition to a vertical tension band to neutralize displacement forces, horizontal fixation helps neutralize rotational forces from the infraspinatus.
The subscapularis inserts onto the lesser tuberosity. These fractures avulse the lesser tuberosity anteromedially. Horizontal fixation best neutralizes these fractures. In four-part fractures, the tuberosities are displaced, and the supportive structures of the articular segment are removed. Therefore, this fragment tilts superiorly and subsides. If the forces then axially load the shaft against this head segment, it can extrude laterally, disrupting the medial calcar and its blood supply.
Neurovascular supply
Some 21-36% of proximal humerus fractures are associated with neurovascular injuries; 8% result in permanent motor loss. The axillary nerve is the nerve most commonly injured. The fracture pattern most commonly associated with axillary nerve injury is an anterior fracture dislocation with a displaced greater tuberosity. Loss of sensation over the lateral deltoid should alert the examiner to possible axillary nerve injury. Isometric contraction of the deltoid should also be tested.
The suprascapular, radial, and musculocutaneous nerves also are at risk. Vascular injuries occur rarely, but 27% of axillary artery injuries may have palpable pulses due to scapular collateral circulation. Associated paresthesias and an enlarging mass must be viewed with caution. Most vascular injuries (84%) occur in patients older than 50 years; 53% are associated with brachial plexus injuries.
Pathophysiology
In attempting to reduce tuberosity fragments, it is vital to take into account the regional differences in the proximal humerus. The cortex of the proximal humerus near the greater tuberosity becomes progressively thicker distally. The exact location of the fracture line depends on the mechanism of, and energy from, the injury.
In fractures in the thinnest cortical bone, the fracture lines can be difficult to appose. These fractures are produced by low-energy forces, occur in porotic bone, and typically are comminuted. Conversely, the denser cortical bone near the biceps groove, and more distally on the shaft, provides an easier surface to approximate fracture lines. Fractures in this area are produced by high-energy forces; the fracture pattern depends on the applied force.
Indirect forces cause most shoulder fractures. The predominant force can cause predictable fracture patterns. Such injury forces are tension, axial compression, torsion, bending, and axial compression with bending. The primary fracture patterns from these forces are transverse, oblique, and spiral.
For each fracture pattern, a preferred method of fixation has been developed to resist displacement forces. Unfortunately, these patterns have not been well described in the shoulder. The orientation of the fracture pattern as a result of tension depends on the muscle-tendon unit that produced most of the displacement force. Treatment recommendations for these fractures are based on factors such as patient motivation, medical history, coexisting medical morbidities, and—the most influential factor—fracture type.
Fracture classification is being reconsidered. Neer's four-part classification, with modifications of the four-part valgus impacted type being separated from four-part fractures in which the humeral head has been extruded laterally, is used primarily to separate these fractures into treatment groups. The majority of fractures are nondisplaced, and nonoperative treatment usually is appropriate. With fracture displacement, operative intervention typically is necessary.
Operative treatment includes closed reduction with percutaneous fixation, open reduction and internal fixation (ORIF), humeral head replacement, and reverse shoulder arthroplasty. [8] Fracture patterns best suited for arthroplasty are as follows:
-
Four-part fractures
-
Fracture dislocations
-
Head-splitting fractures
-
Impaction fractures
-
Humeral head fractures with involvement of more than 50% of the articular surface
-
Three-part fractures in elderly patients with osteoporotic bone
However, heterogeneity of fracture patterns is observed within these groups.
Etiology
The most common mechanism for proximal humerus fractures is a fall on an outstretched hand from a standing height. In younger patients, high-energy trauma is a more frequent cause, and the resultant injury is more devastating. Additional mechanisms include violent muscle contractions from seizure activity, electrical shock, and athletic injuries. Finally, a direct blow to the proximal humerus may also lead to fracture.
Epidemiology
A conservative estimate is that proximal humerus fractures account for approximately 5% of all fractures. These fractures occur primarily in older patients, many of whom are osteoporotic. Like hip fractures, proximal humerus fractures are a major cause of morbidity in the elderly population. As the population base ages, the incidence of these fractures will continue to increase.
Prognosis
The overall prognosis for proximal humerus fractures depends on numerous factors, including the following:
-
Fracture pattern (Neer or Orthopaedic Trauma Association [OTA] type)
-
Patient age
-
Overall health status of patient (associated comorbidities)
-
Patient's expectations
-
Willingness of the patient to undergo lengthy rehabilitation
-
Ability to anatomically reduce the tuberosities in surgically managed fractures
-
Presence or absence of inferomedial support
These fractures as a whole require at least 1 year for recovery.
Yang et al reported functional outcomes following treatment with a proximal humerus locking plate in 64 patients followed for more than 1 year. [9] They divided the study population into two groups according to the presence or absence of inferomedial mechanical support of the humeral head segment.
The authors noted no differences between groups with regard to age, sex, mechanism of injury, or fracture pattern (Neer or OTA). [9] They observed higher Constant-Murley subscores in strength and range of motion in patients with preserved inferomedial support. No difference was noted in pain or activities of daily living portions of the score. They determined that both the presence of an intact medial support and age were independent predictors of functional outcome.
Furthermore, Yang et al noted fracture union in all cases with a mean neck-shaft angle of 126.5° (range, 101-143°). [9] Radiographic differences between those with and without medial support were not reported. Finally, they observed three instances of tuberosity malunion. All occurred in patients older than 65 years with osteopenia and were associated with poor results. The most frequent complication was screw penetration after fracture collapse and loss of reduction. This necessitated screw removal in five of 64 patients.
A randomized controlled trial evaluated the 2-year outcome of locking plate fixation versus nonoperative treatment in elderly patients treated for a displaced three-part fracture of the proximal humerus. The findings report that while treatment with a locking plate resulted in superior functional outcome and health-related quality of life compared with nonoperative treatment, 30% of the patients studied required additional surgery because of fracture complication. [10]
It is important to note that whereas the Constant score, the DASH (Disabilities of the Arm, Shoulder, and Hand) score, and the EQ-5D (EuroQol Group; Rotterdam, Netherlands) score noted in the study were all in favor of the locking plate group on all follow-up occasions, this favorable tendency did not reach statistical significance. [10]
Hatzidakis et al studied the outcomes of 38 patients who were treated with locked angular-stable intramedullary nail fixation for acute two-part surgical-neck fractures at a minimum 12-month follow-up. [11] All fractures healed primarily. The mean Constant score was 71, which was a mean age-adjusted Constant score of 97%. The average forward elevation was 132°. The average Constant pain score was 13 (15 = no pain). In all, 37 (97%) of 38 patients were satisfied with the results. Four patients (11%) required a reoperation.
Südkamp et al evaluated the complication rate and functional outcome of 187 patients after ORIF of proximal humerus fractures using a locking proximal humerus plate. [12] At 12-month follow-up, the average Constant score was 70.6, which was 85% of the score for the contralateral side. The average active elevation was 132°, and external rotation was 45°. The overall complication rate was 34% (52/155), and the most common complication (21/155) was intraoperative screw penetration of the humeral head. Twenty-nine patients (19%) required a reoperation.
Bahrs et al assessed the Constant score and radiographic outcome in 66 patients with minimally displaced and/or impacted fractures of the proximal humerus treated with early immobilization. [13] All of the fractures healed well, without nonunion. In 80% of patients, radiologic assessment showed fracture-displacement of less than 15º angulation and/or less than 5 mm of displacement of the greater tuberosity.
In this study, there was a significant association between the final Constant score and age, American Society of Anesthesiologists (ASA) classification, AO classification, and initial fracture displacement. [13] The authors concluded that early physiotherapy with a short period of immobilization is sufficient management for minimally displaced and/or impacted fractures of the proximal humerus.
Lenarz et al reviewed 30 patients who underwent reverse shoulder arthroplasty for displaced three- and four-part fractures (mean age, 77 years; follow-up, ≥12 months). [14] The mean postoperative American Shoulder and Elbow Surgeons (ASES) Standardized Shoulder Assessment Form score was 78, the mean forward elevation was 139°, and the mean external rotation was 27°. The complication rate was 10%. The authors concluded that reverse shoulder arthroplasty relieved pain and improved function, with a complication rate comparable to those of other treatments.
