Mr Will Eardley
MSc PgCertMedEd DipSEM(UK&I) MD FRCSEd (Tr&Orth)
Consultant Orthopaedic Trauma and Limb Reconstruction Surgeon
Research Associate, University of York CTU | Chairman AOUK Research and Professional Development
South Tees Hospitals NHS Trust
Fracture of the humerus complicated by radial nerve injury is the most common nerve lesion complicating fractures of long bones. The anatomical course of the nerve is critical to understanding the injury, both at the time of fracture and throughout the surgery/repair process.
Ortiginating from the posterior cord of the brachial plexus, the radial nerve takes broad input from cervical roots one through five and also the first thoracic nerve root.
Coursing posterior to the axillary artery on the back wall of the axilla it gives off branches including those of the long and medial head of the triceps. Following passage through the triangular interval the nerve comes in contact with the humerus proper, passing distally closely opposed to the bone in the spiral groove; again supplying both muscle and skin of the upper arm and sensation from the posterior aspect of the forearm.
Transition from the posterior to the anterior compartments of the arm occurs by passage through the lateral intermuscular septum, around seven and a half centimeters proximal to the distal humerus articular surface.
From this point distally towards the elbow joint line, further branches arise to supply forearm musculature and sensation in the radial web space.
When the humerus fractures, damage can occur via two mechanisms.
Firstly, the close apposition of the nerve to the bone through the middle third of the upper arm means it is vulnerable to the sharp bone fragments, or getting caught up within the fracture itself.
Secondly, the nerve is somewhat tethered as it passes through the lateral inter-muscular septum of the arm. Transition away from the bone surface and through into the anterior compartment places the nerve at risk as it transits the septum.
Fractures to the middle and lower thirds of the humerus therefore carry a much stronger association with nerve injury at time of fracture and any subsequent manipulation. The previously over emphasized relationship between the distal third humeral fracture and nerve injury has been found to be erroneous and actually injury of the lower two thirds should be seen as carrying equal risk of nerve insult.
In addition to the opportunity for nerve trauma at time of fracture, iatropathic injury can occur during surgery both from direct trauma and indirect traction or otherwise applied pressure.
Clearly no surgeon would seek to cause nerve trauma. Decision-making is therefore solely around the optimum management of the clinical finding of radial nerve palsy either immediately following first assessment at time of injury, after a closed manipulation or following surgery.
Several authors have endeavored to provide insight into the complex area of how best to navigate a course for patient when radial nerve injury is found following humeral fracture. The nerve may recover. The fracture will most likely heal. The balance is in the risks and benefits of performing surgery to explore to inform, or to address a functional deficit that may/may not definitely recover should surgery not be performed. The difficult issues facing surgeon and patient are obvious.
In common with other areas of fracture care, evidence to guide treatment is often limited to case series and low number controlled trials. Nevertheless an overview of the relevant literature can draw out the key issues to inform decision-making and later case assessment.
Shao et al (2005) reported a systematic review of 35 papers detailing radial nerve injury with humeral fractures 1964-2004. The overall prevalence of radial nerve palsy after fracture of the shaft of the humerus was 11.8% (532 palsies in 4517 fractures) in their review. Fractures of the middle and distal parts of the shaft had a significantly higher association with radial nerve injury than more proximal fractures. Transverse and spiral fractures were seen to be more likely to be associated with nerve injury than oblique or comminuted patterns of fractures. Overall, the rate of recovery was 88.1% (921 of 1045 patients), with spontaneous recovery reaching 70.7% (411 of 581) in patients treated without surgery. Shao et al in particular found no significant difference in the final outcome when comparing groups which were initially managed expectantly with those explored early. This led the authors to conclude that expectant treatment did not affect the extent of nerve recovery adversely, and waiting would avoid many unnecessary operations.
Whilst there is definite value in the discrete findings of prevalence and fracture site location for instance, there are some methodological challenges to the conclusions drawn. When saying expectant treatment is no worse than surgical intervention, it must be appreciated that the authors themselves advised caution when comparing these groups.
In particular, and fundamental to the analysis, Shao et al. combined nonsurgical and late surgical management together when comparing with early exploration. Especially when seen in the light of the globally poor literature from which these findings are drawn, caution should indeed prevail because comparing early exploration to two groups (not explored and late explored) and coming up with the conclusion that there is no difference in outcome is flawed.
The key question is: In the presence of a radial nerve palsy in the setting of a humeral shaft fracture, should the nerve be explored?
Ilyas et al. highlighted the limitations of the work by Shao et al. when advocating for further review of this clinical conundrum. In the face of rates of spontaneous recovery reaching 70%, can early exploration be justified? Providing an update, Ilyas et al. examined 23 articles from 2000 to date in addition to those of Shao et al. This yielded 890 palsies seen in 7,262 patients, giving a similar overall prevalence of 12.3%. Spontaneous recovery (no operation to address the nerve) was seen in 77.2%, similarly corroborating the finding of 70.7% of Shao et al.
This is where the agreement in these two reviews ends. When the impact of early intervention is considered, Ilyas et al reach different conclusions, having analysed the data in perhaps a more meaningful manner.
In this 2020 review, they showed that patients who underwent nerve exploration more than 8 weeks after their injury had a rate of recovery of 68.1%. In contrast, patients treated with early surgical exploration & fracture repair (within 3 weeks of injury) had a rate of recovery of 89.8%.
This suggests that late intervention yields a nerve recovery not dissimilar to expectant treatment, whereas early recovery may improve outcome.
The veil of caution should not be discarded however. The reader is reminded of the often uncontrolled, heterogenous population. This lack of similarity and control within patients is additive to lack of controlled surgical assessment and philosophy. It is impossible to confidently state that early exploration is suggested by the available literature. It is more accurate to say that when it is performed, good results are often obtained.
It can be concluded therefore that in about 12% of humeral shaft fractures, a radial nerve palsy may occur. Of these cases, over 70% will most likely recover spontaneously. It is difficult to be more certain than that, particularly when it comes to exploration and / or its timing.
Shao et al. helped to demonstrate the size of the problem by providing an estimate of cases needed to be enrolled in a randomised controlled trial in order to give level five evidence. 969 patients would be needed for each arm with a total of 1938 patients. Then, if those numbers were able to be recruited, if a difference did exist, there would be an 80% chance of identifying it. It’s important to remember, that even with nearly 2000 randomised cases, there is a 20% risk of an incorrect finding.
In order to decide if a nerve should be explored following isolated, closed injury, informed discussion with the patient should lead to a patient-specific balance of risks and tolerance of complications. How willing is the patient to tolerate uncertainty for several weeks?
There is no doubt that for some patients in whom surgical risks are low and intolerance of delay is high, early exploration and stabilization would be beneficial. For other high-risk or low-demand individuals, the balance is considerably tipped away from exploration towards seeking that 70% spontaneous recovery.
In the meantime however, as in every radial nerve palsy, the full range of motion of distal joints must be maintained from the outset with a physiotherapy regime and the use of dynamic splints.