➢ Most humeral shaft fractures will heal without operative treatment.
➢ When indications for operative treatment are met, plate fixation is reliable and safe.
➢ Nail fixation may be helpful in pathologic or highly comminuted fractures, but use of nails for routine fractures is associated with more shoulder dysfunction than plating or nonoperative treatment.
➢ Radial nerve palsy in the closed fracture setting usually resolves without surgical intervention.
Fractures of the humeral shaft account for 1% to 3% of all fractures and for 20% of all fractures of the humerus1. This fracture has a bimodal distribution; the first spike in prevalence is in young males and is typically associated with high-impact trauma, and the second spike is in elderly women between the ages of sixty and eighty years2. Humeral shaft fractures are also an independent predictor of intra-abdominal, long-bone, and hand fractures in the patient with trauma3. These fractures cause temporary disability in the young and can cause permanent disability in the old4. Reasonable treatment options include nonoperative management, plating, and intramedullary nailing.
Use of Levels of Evidence in the Assessment of Scientific Information
The Levels of Evidence guideline for The Journal of Bone & Joint Surgery was established in January 20035. This guideline provides Levels of Evidence that are based on a hierarchical rating system for the classification of study quality. A five-level rating system was developed for four different study types: therapeutic, prognostic, diagnostic, and economic or decision modeling. Recommendations for or against different aspects of treatment of humeral shaft fractures were graded with use of these guidelines. The reasoning pertaining to each treatment is explained next. A summarized review of this information and details on grades of recommendation are shown in Table I.
Historically, nonoperative treatment has been the mainstay of treatment for isolated humeral shaft fractures. The Orthopaedic Trauma Association (OTA) fracture classification6 distinguishes the humeral shaft from the proximal and distal parts of the humerus by the rule of squares, as with other long bones. The greatest width of the proximal or distal parts of the humerus defines the length of the side of the square. Shortening of humeral fractures is well tolerated functionally. Fracture reduction is greatly assisted by the influence of gravity on the fractured upper arm in the anatomic position. Because of its non-weight-bearing nature and soft-tissue envelope, some deformity can be concealed7. The acceptable alignment for this fracture is 20° of anterior bowing, 30° of varus angulation, 15° of malrotation, and 3 cm of shortening8.
Several studies in the literature have shown high union rates and minimal complications with nonoperative management. The standard of nonoperative treatment is functional bracing. A study performed by Sarmiento et al. found a union rate of 97% (604 of 620)9. One of the many advantages of Sarmiento bracing is that the joints above and below the fracture are not immobilized, allowing range of motion of the extremity, thus resulting in less joint stiffness. Another advantage is the ability to remove the brace for short periods of time to maintain axillary hygiene.
Fracture alignment is maintained through the soft-tissue envelope, which has been termed the “inner splint”10. This inner splint, along with the compression of the soft tissue applied by functional bracing, facilitates healing and alignment. The greater callus strength is attributed to an increase in blood flow to the fracture by the occurrence of physiologically controlled motion4.
Indications for functional bracing have been cited as closed diaphyseal fractures without marked distraction between the fragments, closed fractures associated with initial radial nerve palsy and similar to those described in the previous category, and open fractures without extensive soft-tissue damage9.
Contraindications for nonoperative treatment include polytrauma, obesity, pendulous breasts, and noncompliance. Nonoperative management has the advantages of avoiding operative intervention, cost reduction, no risk of infection, and a decreased risk of iatrogenic nerve palsy. Sarmiento et al. reported on fifty-one humeral fractures and found that union occurred in 98% (fifty fractures), with good functional outcomes. Interestingly, the only nonunion in that study occurred with a pathologic fracture11. Since the publication of the study by Sarmiento et al. and other studies, closed management has become accepted as the standard of treatment for isolated humeral shaft fractures.
Closed treatment is still in debate because of the lack of reproducibility of these results. Some studies have found a much higher nonunion rate with nonoperative management, especially in comparison with an operative group1. A retrospective study on 213 patients found that 21% (thirteen) of the sixty-three patients treated nonoperatively and 9% (thirteen) of the 150 patients treated operatively went on to nonunion12. The reasons for the variable union rates have included variable methods used to measure time to union, patients lost to follow-up, and patient or surgeon compliance with bracing. For functional bracing to work optimally, the patient must continue to maintain joint motion and must continually adjust the functional brace to keep adequate compression.
Nonunion Due to Fracture Location
The prevalence of nonunion in specific locations and fracture patterns of the humeral shaft may be due to various factors. A retrospective case series found that oblique fractures had a high nonunion rate of 11% (four of thirty-six fractures) and attributed this to soft-tissue interposition and deforming muscle forces at the fracture site. Simple transverse patterns also had higher rates of nonunion (8% [one of twelve fractures]) that were attributed to the scant surface area for bone-to-bone apposition and the greater amount of fracture site strain. With regard to fracture location, proximal-third fractures had a nonunion rate of 29% (four of fourteen fractures) compared with 5% (one of twenty-two fractures) in the middle third and none in the distal third. That review concluded that proximal oblique fractures had the highest rate of nonunion, whereas middle-third oblique fractures had the highest rate of union, when treated nonoperatively13.
Another descriptive study found that simple fractures (type A according to the OTA classification) had a higher nonunion rate (18% [seven of thirty-nine fractures]) compared with type-B fractures (4% [one of twenty-six fractures]) and type-C fractures (0% [zero of thirteen fractures]) in the proximal and middle third of the shaft when treated nonoperatively14. A comparative cohort study found that even extra-articular distal-third fractures of the humerus that might theoretically be difficult to control with closed treatment achieved similar results when treated with operative or nonoperative management15.
Indications for Operative Management
Indications for operative management of humeral shaft fractures include open fractures, high-energy fractures associated with brachial artery injury, segmental fractures, fractures combined with radial and ulnar fractures, pathologic fractures, and the need for early mobilization in patients with polytrauma and associated injuries including lower-extremity trauma, which requires weight-bearing through the upper extremity for crutch and/or walker use. High-energy trauma is classified as involving open fractures with severe soft-tissue injury, multifragment fractures, wounds from bullets with a projectile velocity of ≥1200 m/s, and highly comminuted fractures16. Relative indications include obese patients who do not tolerate bracing well and noncompliant patients. Associated brachial plexus injury is also a relative operative indication, as it is thought that the loss of brachial muscle tone and contractions lead to more malunions and nonunions with nonoperative treatment1.
Methods of Operative Management
Intramedullary nailing is a method of operative treatment for humeral shaft fractures. Good results were initially reported. A Medicare cohort study on 511 patients undergoing operative treatment was performed over a period of fifteen years from 1993 through 2007. Of these 511 patients, 39% (201 patients) underwent plate fixation and 61% (310 patients) underwent intramedullary nailing. That study concluded that intramedullary nailing was performed more frequently than open reduction and internal fixation17.
In the early years of humeral nailing, the Seidel nail initially was reported as having good results but was associated with several complications. These complications included rotational instability, iatrogenic fracture, and severe shoulder injury caused by insecure fixation and a large nail head. The later Russell-Taylor nail with transfixing screws had a smaller nail diameter and better fixation. The complications of the Russell-Taylor nail included rotational instability caused by a discrepancy between the size of the screw and the screw hole. The more recent Synthes nail has had better success clinically with fewer complications18.
Indications for intramedullary nailing are fracture comminution and segmental and pathologic fractures19. Benefits of intramedullary nailing of humeral fractures include a less invasive procedure than open reduction and plate fixation, autograft material produced from reaming, the biomechanical properties of load sharing, and higher moments of inertia20. Physiologically, nailing allows for the preservation of the periosteal blood supply and minimizes the disruption of fracture biology. Lower infection rates and lower incidence of iatrogenic injury to the radial nerve have been reported as well21.
Historically, there has been a higher rate of complications and shoulder dysfunction associated with intramedullary nailing. A prospective randomized trial that compared intramedullary nailing with dynamic compression plating indicated that there were more complications in the intramedullary nailing group (containing twenty-one patients) that required a second surgery in comparison with the dynamic compression plating group (twenty-three patients). In that study, 33% (seven patients) in the intramedullary nailing group required reoperation, and 4% (one patient) required reoperation in the dynamic compression plating group. Shoulder impingement occurred in 29% (six patients) in the intramedullary nailing group but in only 4% (one patient) in the dynamic compression plating group22. Another prospective study confirmed these results. Complications were found to be more frequent in the intramedullary nailing group at 58% (twenty-two of thirty-eight patients) in comparison with the dynamic compression plating group at 43% (twenty of forty-six patients). With regard to shoulder range of motion, there was a decreased range of motion of the ipsilateral shoulder (>10° loss of abduction, forward flexion, or external rotation compared with the contralateral side) in 16% (six patients) in the intramedullary nailing group and a decreased range of motion (>10° loss) in 13% (six patients) in the dynamic compression plating group. Of note, all patients in the dynamic compression plating group with restricted elbow motion had fractures at the junction between the middle and distal thirds or in the distal third of the humerus. In all of these cases, the distal end of the plate was located just proximal to the olecranon fossa22. The Cochrane review that investigated intramedullary nailing and dynamic compression plating found that dynamic compression plating was associated with a decreased risk of shoulder impingement (relative risk of 0.12) and the need for implant removal (relative risk of 0.17)23.
There is very little in the literature that provides insight into which fracture patterns are better treated with intramedullary nailing compared with dynamic compression plating.
Types of Intramedullary Nails
Although there are potentially many different kinds of intramedullary nails for the humeral shaft, the most common are rigid, reamed, locked nails. These nails can be placed antegrade or retrograde, with antegrade being by far the most commonly used technique. There is no general consensus on when to use which approach. Some studies have suggested that antegrade nailing be performed on proximal and middle-third shaft fractures and retrograde nailing be performed on distal-third fractures1. Another study has suggested the use of antegrade nailing with fractures of the proximal half and the use of the retrograde approach for the distal half23. Retrograde nailing has the benefit of sparing shoulder pain and dysfunction. Conversely, the disadvantages to this technique are the increased incidence of elbow pain postoperatively and the increased risk of fracture extension to the humeral condyles. Union rates are equal between these two procedures. A prospective comparison of midshaft fractures treated with antegrade and retrograde intramedullary nailing found no significant difference in union rate, which was 95% (forty-two of forty-four) in the antegrade group and 93% (forty-two of forty-five) in the retrograde group.
Although nonunion rates are the same for these types of treatment, the causes are different. A prospective, randomized comparative study stated that fracture distraction during antegrade nailing is thought to contribute to nonunion and that longer nails come with a higher risk of distraction. Fracture comminution is greater with retrograde nailing as the hoop stresses are high on intramedullary nail insertion and retrograde intramedullary nailing has an increased risk of fracture propagation18.
Plating has provided good results, with a high rate of union and good function with a fairly low rate of complications. Most literature is consistent in showing good outcomes with this method of treatment. A retrospective review on forty-four humeral shaft fractures found that 97% (forty-three of forty-four fractures) went on to union after the initial operation. Shoulder and elbow function essentially returned to normal24.
Another study found that 2% (two) of 102 fractures went on to nonunion. Functional outcomes in that study were consistent with the literature, with 87% (eighty-nine of 102 patients) having good or excellent functional recovery at the shoulder and elbow25.
Another study investigated open reduction and internal fixation of the humerus in thirty-five cases and found that nonunion occurred in 6% (two cases). Iatrogenic radial nerve palsy occurred in 9% (three cases). Functional outcome was consistent with the literature, with 74% (twenty-six cases) showing full range of motion and 94% (thirty-three cases) showing full strength26. A randomized controlled trial comparing nailing with dynamic compression plating found that nonunion was more frequent with plating at 6% (one of eighteen) compared with 0% (zero of eighteen) in the nailing group27.
Traditionally, external fixation has been the mode of treatment for open fractures of the humerus. Some surgeons use dynamic compression plating in the acute setting, with very promising results. A retrospective review on forty-six patients who underwent immediate open reduction and internal fixation of open humeral fractures found that immediate plate osteosynthesis for such fractures yielded excellent results. With regard to bone union, all fractures in that study united without any need for reoperation. Thirteen percent (six of forty-six fractures) had the complication of delayed union. Interestingly, there were no deep infections reported28.
Weight-Bearing of the Humerus After Operative Treatment
Early mobilization was thought to be safer for patients who underwent nailing instead of plating of the humerus. This concept has been postulated because of the successes of early weight-bearing with nails in the lower extremity. The literature on the safety of weight-bearing after humeral shaft fractures and its effect on union rate is not robust for conventional plating. A retrospective study that investigated the union rate with and without weight-bearing performed a statistical analysis with use of the Fisher exact test and found no significant effect with regard to weight-bearing. In the study, 94% (seventy-eight of eighty-three fractures) achieved union after the initial surgery; 6% (two of thirty-three fractures) in the non-weight-bearing group and 6% (three of fifty fractures) in the weight-bearing group required a second surgery to achieve union. That study also found that radiographic alignment did not differ between the groups29.
Dynamic compression plating is used for treatment of most humeral shaft fractures. In some patients, the use of locked plates is indicated. These indications are osteoporotic bone, very proximal or distal fractures on the humeral shaft, and highly comminuted fractures1. Recently, there has been an increase in usage of small fragment plates as a reduction tool in fractures where reduction cannot be maintained during the application of a conventional dynamic compression plate. Some surgeons will also use these small fragment plates on patients of small stature30.
The operative approach used for the treatment of humeral shaft fractures is dictated by the location of the fracture. The anterolateral approach is used for proximal diaphyseal fractures. The structures at risk during this approach are the lateral antebrachial cutaneous nerve and the radial nerve distally. The radial nerve is protected proximally by the lateral portion of the brachialis muscle. The posterior approach is an extensile approach that is typically used for fractures of the distal third of the humeral shaft. This approach allows for the visualization of the radial nerve in the spiral groove, and the nerve can be mobilized for more proximal fractures. The main structure at risk here is the radial nerve31. The lateral approach can be used for extensile injuries either proximally or distally. One advantage to this approach is that the patient is in the supine position. Another advantage is that the distal two-thirds of the humerus can be exposed through the intermuscular plane between the triceps muscle and anterior arm musculature. The structures at risk during this approach are the radial nerve and posterior antebrachial cutaneous nerve. The potential risk of iatrogenic radial nerve injury varies among different approaches. A retrospective study that compared the posterior approach and the lateral approach found that the posterior approach had a 12% incidence (three of twenty-six cases) of iatrogenic radial nerve palsy and the lateral approach had a 0% incidence (zero of thirty cases) of radial nerve palsy32.
Minimally Invasive Plating
Minimally invasive plating is an innovative technique that is currently not widely performed. One of the disadvantages of conventional open reduction and internal fixation is the amount of soft-tissue dissection needed to achieve adequate reduction and stabilization. Minimally invasive plate osteosynthesis has been developed for anterior plate and screw stabilization with less soft-tissue disruption. Theoretically, this improves the healing rate while reducing complications such as infection and iatrogenic radial nerve palsy. One of the major difficulties of minimally invasive plate osteosynthesis is adequate fracture reduction. Reduction in conventional open reduction and internal fixation is through direct visualization, and minimally invasive plate osteosynthesis requires indirect reduction and closed fracture manipulation.
Studies of minimally invasive plate osteosynthesis have consistently shown union rates of >90% even in the setting of open fractures. A study of thirty-five fractures found that at twelve weeks, 91% (thirty-two fractures) achieved union. The mean active range of elbow motion in that study was 114°. Only 57% (twenty fractures) achieved full elbow extension, 31% (eleven fractures) had an extension lag of 20°, 6% (two fractures) lost 15° of extension, and 6% (two fractures) lost 30° of extension33. Another retrospective study investigated conventional plating and minimally invasive plate osteosynthesis; it found that 6% (one of sixteen) in the conventional plating group went on to nonunion, and 0% (zero of seventeen) did so in the minimally invasive plate osteosynthesis group. Iatrogenic radial nerve palsy did not occur in the minimally invasive plate osteosynthesis group but did occur in 31% (five of sixteen) in the conventional open reduction and internal fixation group. With regard to function, the mean University of California Los Angeles (UCLA)34 end-result score was 34.76 points in the minimally invasive plate osteosynthesis group and 34.38 points in the open reduction and internal fixation group35. Another retrospective cohort study mirrored these findings, with 10% (three of thirty patients) in the open reduction and internal fixation group going on to nonunion and 3% (one of twenty-nine patients) in the minimally invasive plate osteosynthesis group going on to nonunion. The mean UCLA shoulder scores were not significantly different, with 33.8 points in the open reduction and internal fixation group and 34.3 points in the minimally invasive plate osteosynthesis group. The mean Mayo36 elbow scores in this study were 97 points in the open reduction and internal fixation group and 97.6 points in the minimally invasive plate osteosynthesis group37.
Dynamic Compression Plating Compared with Intramedullary Nails
There are several studies that have investigated the outcomes of dynamic compression plating and intramedullary nailing12,17,19,20,22,23,27,38-40. The benefits of open reduction and internal fixation include direct visualization of the fracture, anatomic fracture reduction, interfragmentary compression, potential identification and protection of the radial nerve, and no violation of the adjacent joints21. The disadvantages associated with plating are the increased soft-tissue stripping, risk of radial nerve injury, risk of fracture at the end of the plate, and poor screw purchase in osteoporotic or comminuted bone41. An increased incidence of elbow stiffness has been seen in plating compared with intramedullary nailing of humeral shaft fractures as well.
Shoulder dysfunction after antegrade intramedullary nailing has been cited because of antegrade nail insertion near the rotator cuff. A retrospective cohort study of forty-eight patients examining the consequences of the anterolateral approach for antegrade intramedullary nailing on the rotator cuff using the Constant Score, Simple Shoulder Test, and shoulder examination indicated that 79% (thirty-eight patients) obtained a good result42. Only 2% (one patient) had a limited functional outcome. Sonographic assessment found that there was a partial supraspinatus tendon tear in 6% (three patients) and only 3% (one patient) had a complete rupture. The authors concluded that this approach offered a good functional outcome with little impact on the tendon42.
In contrast, intramedullary nailing has repeatedly been shown in other studies to have a high complication rate at the site of nail entry. If antegrade nailing is used, the shoulder seems to have lower function scores. If retrograde nailing is performed, the elbow seems to sustain increased pain and decreased range of motion. A study that investigated the long-term results of shoulder and elbow function after humeral shaft fracture treatment was performed on 128 patients. These patients were separated into three treatment groups: the first group received functional bracing, the second group received dynamic compression plating, and the third group received intramedullary nailing. The mean Constant score was highest in the functional bracing group at 92.4 points, followed by the plating group at 85.6 points and the intramedullary nailing group, with the lowest score, at 74 points43. A retrospective study of twenty-nine patients treated with plating and forty-four patients with antegrade intramedullary nailing found that all of the parameters measured were significantly lower (p = 0.001) on the injured side in comparison with the contralateral noninjured side in both groups. The L’Insalata score44 was 83 points (88% of the noninjured side) in the dynamic compression plating group and 82 points (91% of the noninjured side) in the nailing group. The mean visual analog scale (VAS) for the pain portion of the Constant score was 12 points for the dynamic compression plating group and 10 points for the nailing group. That study concluded that shoulder joint range of motion and strength do not recover to normal after humeral shaft fracture and that antegrade intramedullary nailing, if performed properly, is not responsible for shoulder joint impairment45. Malrotation of the humerus has been implicated as a cause of the loss of shoulder range of motion after nailing of the humerus. A randomized controlled study investigating malrotation postoperatively with computed tomography (CT) scans indicated a significant difference (p = 0.05) between intramedullary nailing and plating. Internal rotation of ≥20° was seen in 27% (six of twenty-two patients) in the intramedullary nailing group in comparison with 0% in the plating group. This malrotation directly correlated with the amount of shoulder dysfunction that was reported46. With regard to reoperation rate, a meta-analysis that was performed found that plate fixation had a lower relative risk of reoperation of 0.26 (95% confidence interval [CI], 0.007 to 0.9; p = 0.03) compared with intramedullary nailing. This translates into a risk reduction of 74% when plate fixation was the method of treatment38. Some studies found that the complication rate was similar between dynamic compression plating and nailing. A meta-analysis of 203 patients from four studies found that there was no significant difference in complications between plating and nailing (relative risk, 0.56 [95% CI, 0.30 to 1.04]; p = 0.07)39.
These identified complications have shown that intramedullary nailing, regardless of the starting point, is associated with higher technique-related complications in comparison with open plating techniques. A meta-analysis of randomized controlled trials and nonrandomized studies comparing dynamic compression plating and locked intramedullary nailing concluded that there was a higher risk of technique-related complications and impairment of shoulder function in the intramedullary nailing group compared with the dynamic compression plating group40. A review of the complications is summarized in Table II for intramedullary nailing and in Table III for dynamic compression plating.
Radial Nerve Palsy
Radial nerve palsy is associated with humeral shaft fractures, either as a result of the injury or during treatment. According to a systematic review, the overall prevalence of primary radial nerve palsy after humeral shaft fractures is 12% (532 of 4512 fractures)47. The incidence does vary with fracture pattern and fracture location. A retrospective study of twenty-four cases of radial nerve palsies found that 58% (fourteen cases) were a result of fractures of the middle third of the humerus, followed by distal-third fractures at 21% (five cases)48. Another retrospective review of thirty-eight cases of radial nerve palsies mirrored these findings, with 58% (twenty-two cases) occurring in the middle third of the shaft, 13% (five cases) occurring in the distal third of the shaft, 11% (four cases) occurring in the proximal third of the shaft, and 16% (seven cases) not specified49.
A systematic review analyzing radial nerve injury suggests that transverse fractures (21% [forty-seven of 222]) and spiral fractures (20% [nineteen of ninety-six]) are more likely to be associated with radial nerve injury, but oblique patterns (8% [fifteen of 179]) or comminuted patterns (7% [twenty-six of 382]) were least likely47. The highest incidence of radial nerve palsy occurs with a spiral fracture in the distal third of the humerus. The frequency of radial nerve palsy was greater with this Holstein-Lewis fracture pattern at 22% (six of twenty-seven) compared with other fracture patterns at 8% (twenty-seven of 334)14. The nerve lies close to the inferior lip of the spiral groove, but not in the groove. It is only near the lateral supracondylar ridge that the nerve is in direct contact with the humerus, and, for that reason, Holstein and Lewis believed that the radial nerve has a higher prevalence of injury in middle and distal-third fractures10. The Holstein-Lewis fracture occurs in the distal third at a point where the radial nerve comes through the lateral intermuscular septum and is in contact with the bone. The force of the injury causes the proximal fragment to displace distally. The intermuscular septum and the radial nerve that is contained within the septum displace distally as well. While this occurs, the apex of the distal fragment moves proximally and radially. This causes the radial nerve to be lacerated or to be trapped in between the two fragments24. There are some studies that show that the zone of injury that is at high risk for radial nerve palsy extends beyond the Holstein-Lewis fracture location. A systematic review that analyzed the location of radial nerve palsies associated with humeral fractures found that if the humerus was divided into five parts, the region that is most at risk appeared to extend from the midshaft to the level of the typical Holstein-Lewis location. The radial nerve palsy rate was 3% (three of eighty-nine) in the most proximal aspect, 11% (six of fifty-seven) in the middle-proximal shaft, 22% (forty-eight of 219) in the middle shaft, 20% (thirty-two of 160) in the middle-distal shaft, and 11% (eight of seventy-six) in the most distal aspect47. With regard to treatment of the Holstein-Lewis pattern, there is debate on whether or not early exploration for a radial nerve palsy is warranted because of the likelihood of nerve entrapment that occurs with this fracture pattern. A retrospective study of sixty-two fractures associated with radial nerve palsy found that of the 18% (eleven) of Holstein-Lewis fractures with radial nerve palsy, 82% (nine fractures) were treated conservatively and 18% (two fractures) were treated with early exploration. All that were treated conservatively fully recovered. In one of the two that were treated with exploration, the radial nerve was caught between the fracture fragments. The patient treated operatively who had no nerve entrapment was lost to follow-up, but the other patient fully recovered50.
Treatment for Radial Nerve Palsy
A retrospective review on the comparison of nonoperative and operative treatment in 186 patients concluded that the rate of radial nerve palsy was not significantly different between the two groups. The rate of radial nerve palsy was 9% (eight of ninety-one patients) in the nonoperative group and 10% (nine of ninety-five patients) in the operative group. Interestingly, four of nine patients in the operative group with radial nerve palsy also sustained iatrogenic nerve injury51.
Open reduction and internal fixation with plates and screws has traditionally been thought to have a higher risk of radial nerve palsy in comparison with intramedullary nailing. A meta-analysis of randomized controlled trials and nonrandomized studies comparing dynamic compression plating and locked intramedullary nailing found that dynamic compression plating had a significantly higher risk of radial nerve palsy in comparison with intramedullary nailing (relative risk of 1.82 [95% CI, 1.02 to 3.26]; p for heterogeneity = 0.54)39.
It is believed that radial nerve palsy that occurs with intramedullary nailing is typically a result of fracture manipulation and should be treated with observation. The one exception to this is with external-rotation spiral fractures, because the etiology of the radial nerve palsy may be nerve entrapment41.
A difficult question is when to perform exploration for a radial nerve palsy. It is understood that if open reduction and internal fixation is to be performed or there is the presence of radial nerve palsy in the setting of an open injury, then exploration of the nerve is indicated. Radial nerve lacerations are typically seen in near-amputations and extensive upper-extremity injury. In a retrospective study, radial nerve exploration was performed on all eleven of the open humeral fractures with radial nerve palsy. Of these eleven open humeral fractures, 55% (six fractures) had radial nerve lacerations. Interestingly, all of the nerves in this study that were primarily repaired failed to recover52.
With closed injuries, the advantages of early exploration are as follows: the procedure is technically easier and safer, direct visualization of the injured nerve could determine whether or not the nerve palsy is due to laceration, reduction of the fracture reduces the risk of further neural damage via mobile bone ends, shortening the humerus can facilitate nerve repair, and early stabilization reduces the chance of nerve entrapment by scar tissue and callus. Some surgeons believe that a period of waiting is appropriate prior to nerve exploration25. Furthermore, by waiting, the level of the nerve injury becomes clearly demarcated because of degeneration of the injured portion of the nerve, allowing the nerve to be cut back to healthy fascicles and leading to appropriate grafting of the defect. Most importantly, there is a high rate of recovery with this type of management. A retrospective study found that it is unusual to see a complete laceration of the radial nerve in closed injuries even if the mechanism is high-energy trauma and it is rare that the radial nerve will not recover in the setting of a closed injury52.
The duration of observation and the method of testing are controversial. In a retrospective study, the mean time to initial recovery was seven weeks (range, one to twenty-five weeks). Another study suggested that nerve exploration should occur within eight to sixteen weeks if there is no return of nerve function53. Some use serial electromyograms to monitor for nerve function. These electromyogram findings have a limited clinical importance because such findings typically parallel clinical findings and show signs of recovery no earlier than one month prior to the clinical evidence.
The authors of one study stated that, when performing a radial nerve exploration, the lateral approach should be used because of the access to the proximal and distal aspects of the nerve52. When performing an exploration, typically osseous landmarks are used to identify the location of the radial nerve. These landmarks usually are not accurate intraoperatively because of fracture distortion and displacement. It has been proposed to use the apex of the triceps aponeurosis as a landmark. A study that was performed on cadavers and sixty patients measured the distance of the radial nerve from this apex. Those results indicate that the distance of the radial nerve was consistently about 2.5 cm from the apex of the triceps aponeurosis54.
Recovery Rates for Radial Nerve Palsy in Nonoperative Management
As discussed above, radial nerve palsy in the setting of a closed injury often resolves without intervention. A systematic review found that 71% (411 of 581 radial nerve palsies) in their analysis resolved with nonoperative treatment. Of the 170 that did not resolve, 87% (149 radial nerve palsies) were treated with further exploration. Nerve exploration consisted of release of any scar tissue impinging on the nerve as well as repair of the nerve if a laceration was present. Of the 149 radial nerve palsies, 66% (ninety-eight) resolved after exploration. The overall rate of recovery was 88% (509 of 581 radial nerve palsies). In this same review, 82% (327 of 397 palsies) that were initially explored later resolved47. The studies above indicated that there is no significant difference in overall recovery rates when comparing groups that were initially managed with observation with groups that underwent exploration earlier. Also, initial expectant management did not affect the extent of radial nerve recovery, indicating that the initial observation of closed injuries would avoid many unnecessary operations52. If no function recurs, a late tendon transfer gives reliable hand function55.
The majority of humeral shaft fractures heal with nonoperative treatment with low risk and minimal dysfunction. Certain fracture locations (proximal) and conditions (brachial plexus palsy) have a higher incidence of nonunion with nonoperative treatment. Open plating is reliable and safe if surgical indications are met. Antegrade nailing may be the best option for pathologic, severely comminuted, or segmental fractures, but is associated with shoulder problems. Exploration of the radial nerve in the presence of palsy associated with the injury should be delayed for several months because spontaneous recovery is common.
Source of Funding: No external funding was utilized in this review.
Investigation performed at the Vanderbilt University Medical Center, Nashville, Tennessee
Disclosure: None of the authors received payments or services, either directly or indirectly (i.e., via his or her institution), from a third party in support of any aspect of this work. One or more of the authors, or his or her institution, has had a financial relationship, in the thirty-six months prior to submission of this work, with an entity in the biomedical arena that could be perceived to influence or have the potential to influence what is written in this work. No author has had any other relationships, or has engaged in any other activities, that could be perceived to influence or have the potential to influence what is written in this work. The complete Disclosures of Potential Conflicts of Interest submitted by authors are always provided with the online version of the article.
- Copyright © 2015 by The Journal of Bone and Joint Surgery, Incorporated