Peripheral nerve injuries in haemophilia

Introduction

The most common form of nerve damage in haemophilia is peripheral neuropathy (peripheral nerve injury), which mainly affects the upper and lower limbs. The purpose of this review is to provide information on the aspects of haemophilia management which may have an influence on or relationship with peripheral nerve injuries (replacement therapy, prophylactic treatment, implications of this complication in patients with inhibitors) and underline the specific features of peripheral neuropathy in haemophilic individuals.

Peripheral nerve injuries in haemophilia

Peripheral nerves injuries are uncommon in patients with haemophilia. However, 20 cases of peripheral nerve involvement in haemophiliacs were reported by Saraf et al¹. Among the 134 patients with bleeding disorders examined, 20 (15%) had features of nerve palsy. The femoral nerve was involved in the majority of cases (n=15), followed by the sciatic nerve (n=4) and the peroneal nerve (n=1). Femoral nerve involvement was associated with iliac bleeding in 13 patients and inguinal bleeding in two cases. All four cases of sciatic nerve palsy were associated with gluteal haematoma. The single case of peroneal nerve palsy was associated with a haematoma in the upper leg and calf.

Lütschg and Vassella² reported on ten patients with peripheral nerve lesions, palsy of the femoral nerve being the most frequent (5 cases). Two patients had a lesion of the lumbar and sacral plexus, two patients a lesion of the radial nerve and one patient a lesion of the cutaneous femoris lateralis nerve. Schmitz et al. reported on a patient with severe haemophilia who presented with acute onset of neck pain with cervical nerve root irritation, due to a spinal epidural haematoma³.

Ulnar and median nerves

Mortazavi et al.⁴ reported on six patients who were diagnosed with tardy ulnar nerve palsy. All patients suffered from severe haemophilia A. Anterior subcutaneous transposition of the ulnar nerve was performed in all except one. No post-operative complications or recurrences occurred. No additional surgery was required in operated patients. After surgery, subjective sensory and motor disturbances were improved or resolved in all of the operated patients. Anterior subcutaneous transposition of the ulnar nerve is an effective procedure, with a low risk of complications.

In haemophilia the mechanism involved in ulnar nerve neuropathy is severe cubitus valgus secondary to advanced haemophilic arthropathy of the elbow joint. In such circumstances the nerve is stretched and the intraneural blood supply diminished. The rationale for decompression or anterior transposition of the ulnar nerve is to eliminate the aforementioned nerve elongation and its complication (tardy ulnar nerve palsy).

Kaymak et al. reported on a boy with haemophilia who suffered a pronounced median neuropathy and a minor ulnar neuropathy due to compartment syndrome of the forearm⁵.

Median nerve (carpal tunnel syndrome)

Two cases of carpal tunnel syndrome in haemophiliacs have been reported by Moneim and Gribble⁶. In one patient, replacement therapy was successful in relieving the condition. However, in the other patient, decompression of the carpal tunnel and internal neurolysis of the median nerve were carried out after replacement therapy had failed. Surgical release of the carpal canal together with the epineurectomy and internal neurolysis resulted in complete recovery. The mechanism underlying carpal tunnel syndrome in haemophiliacs is the same as that in the general population. Carpal tunnel syndrome is more common in patients who suffer from advanced renal disease, gout, diabetes and rheumatoid arthritis.

Carpal tunnel decompression is a very effective operation. Most patients are satisfied with the outcome, although some of them can develop complications. New endoscopic techniques are claimed to reduce the incidence of these complications, although the new techniques are not without complications of their own⁷.

A review of the literature over the past 20 years performed by Molitor and Wimperis revealed only nine cases of carpal tunnel syndrome in haemophiliacs⁸. They suggested that management should be conservative in the first instance, with factor VIII replacement continued for 3 to 5 days. Surgical decompression is indicated if symptoms fail to resolve in the first 24 hours.

Ulnar proper digital nerve of the thumb

Lokey et al. reported a case of traumatic false aneurysm following a puncture wound in the hand of a person with haemophilia, despite factor VIII replacement⁹. Following surgical repair of the false aneurysm and the ulnar proper digital nerve of the thumb 3 weeks post-injury the patient had an uncomplicated recovery.

Femoral nerve

Iliopsoas haematomas typically occur in patients with haemophilia or von Willebrand’s disease^10,11. The volume of these bleeds is often large and the haematoma can cause femoral nerve neuropathy. The typical presentation is a sudden onset of increasing pain in the groin and hip radiating to the anterior thigh, associated with weakened hip flexion. In clinical practice, iliopsoas haematoma should be included in the differential diagnosis of pain localized to the groin-hip area; other diagnoses in this context include osteonecrosis of the hip, spinal complications or even appendicitis when the haematoma is on the right side. Computed tomography scan and/or magnetic resonance imaging (MRI) and ultrasonography of the hip are required not only to confirm the diagnosis but also to monitor the size of the haematoma until its complete disappearance. Absolute bed rest and long-term factor replacement are the mainstays of therapy. On-demand treatment must be enhanced immediately after the diagnosis. Treatment must then be continued for weeks by means of secondary prophylaxis until full disappearance of the haematoma. If sufficient haemostasis is achieved until resorption of the haematoma, this conservative treatment is effective. If, however, the haemostasis is not sufficient, recurrent bleeding into the haematoma may take place leading to the development of a haemophilic pseudotumour in this area. Early and effective factor replacement therapy is essential in the prevention of this complication.

In a review of 297 patients with bleeding disorders, Ashrani et al. found 46 episodes of iliopsoas haemorrhage in 31 patients¹². The mean duration of haematological treatment was 19 days, and the mean time spent in hospital was 12 days. Patients with inhibitors stayed in hospital longer than patients without inhibitors (19 days vs 8 days) and had higher factor consumption, although the total duration of therapy was not significantly different. The rate of recurrent bleeding was 3%.

In a review of 146 haemophiliacs, Balkan et al. found 14 iliopsoas bleeding episodes in eight haemophiliacs¹³. Two patients had a high titre inhibitor against factor VIII and accounted for three bleeding episodes (21%). Iliopsoas haematomas were confirmed by ultrasonography in all patients. The mean duration of therapy with clotting factor concentrate was 8 days. The mean duration of hospitalisation was 5 days. All patients started to receive a physical therapy programme 6 days after the initiation of haemostatic therapy which lasted 3 weeks. Ultrasonographic findings related to iliopsoas haematoma disappeared in all patients within 3 months of the initial episodes.

Iliopsoas haematoma has not been described previously as a complication of posterior spinal decompression until Robinson et al. reported the case of an iliopsoas haematoma with femoral neuropathy appearing 8 weeks after a posterior spinal decompression procedure¹⁴. The compartment was fully decompressed with resolution of the nerve root symptoms within 48 hours.

Peroneal nerve

Traumatic neuropathies of the common peroneal nerve at the head of the fibula are relatively common, but nerve entrapment at this area is rarely encountered. Large et al. described a man with haemophilia in whom common peroneal nerve entrapment occurred secondary to a muscle haematoma¹⁵.

Peroneal nerve palsy is a known adverse effect of total knee arthroplasty (TKA) in the general population. Asp and Rand reported 26 cases of peroneal palsy following 8998 TKA interventions (rate, 0.3%)¹⁶. Eighteen patients had complete and eight had incomplete peroneal palsies. Twenty-three had both motor and sensory deficits, and three had only motor deficits. At an average of 5 years after TKA, recovery was complete for 13 palsies and partial for 12. Complete recovery was more likely in those palsies that were incomplete initially.

Krackow et al. reported the cases of five non-haemophiliacs who were treated by operative exploration and decompression of the peroneal nerve for peroneal nerve palsy complicating TKA¹⁷. None of the patients had had an improvement in peroneal nerve function during extended conservative care. The surgical decompression was performed 5 to 45 months after the index TKA and was followed by improved nerve function in all patients: four of the five patients had full peroneal nerve recovery. All patients were able to discontinue use of their ankle-foot orthoses. These authors concluded that when conservative non-operative measures do not lead to sufficient improvement in nerve function, consideration may be given to operative decompression of the peroneal nerve.

In a report by Schinski et al., 19 non-haemophiliacs with a neurological complication were identified after a retrospective review of 1,476 primary TKA, for an overall incidence of 1.3%¹⁸. Contrary to previously published data, valgus deformity, flexion contracture, the use of postoperative epidural anaesthesia, the prolonged use of pneumatic tourniquets, and pre-existing neuropathy were not associated with the development of peripheral neuropathy after TKA. Conservative treatment was initiated in each of the patients immediately after discovery of the nerve palsy. All patients had at least a partial recovery at the end of follow-up, with most experiencing a complete recovery from symptoms.

In haemophilia, Augereau et al.¹⁹ reported a 7% rate of peroneal nerve palsy after TKA (1 out of 14), and Vastel et al.²⁰ a rate of 3.5% (1 out of 29). Therefore, the incidence of peroneal nerve palsy as a complication of TKA seems to be higher in haemophilic patients (3.5–7%) than in the general population (0.3–1.3%).

Diagnosis

Peripheral nerve injuries can cause symptoms of pain, dysaesthesia and partial or complete loss of sensory and motor function. A thorough clinical history, physical examination, electrodiagnostic evaluation, and relevant radiographic studies should be performed to distinguish a peripheral nerve problem from one involving the spinal cord, brain, bone, or soft tissues. In addition, early neurosurgical consultation should be obtained.

The strength of individual muscles or muscle groups must be graded. A sensory examination must be performed, including testing for light touch, pinprick, two-point discrimination, vibration, and proprioception. The presence of Tinel’s sign is useful for localising a nerve injury. This sign refers to paraesthesias elicited by tapping along the course of a nerve. Progressive distal advancement of Tinel’s sign over time can be useful clinically to follow the course of regenerating sensory axons. However, the presence of Tinel’s sign does not guarantee motor recovery. Return of sweating in an autonomic zone signifies sympathetic nerve fibre regeneration. Reflex changes are also sensitive and early indicators of nerve damage.

Both electromyography (EMG) and nerve conduction studies are useful for distinguishing an upper from a lower motor neuron disorder as well as for diagnosing a primary muscle disease. An understanding of the functional anatomy of the peripheral nervous system can often enable the clinician to localise nerve injuries and lesions with a high degree of accuracy.

Imaging techniques such as X-rays, computed tomography and, most recently, MRI can be valuable diagnostic tools in evaluating peripheral nerve injuries. Fractures can be associated with nerve injuries. Computed tomography is able to delineate soft tissue mass lesions such as tumours or pseudotumours. MRI has proven to be much more effective in resolving the fine anatomical detail of soft tissues. Using conventional and enhanced MRI techniques, it has been possible to visualise both normal and abnormal peripheral nerve structures. New techniques such as MRI “neurography” make it possible to image and reconstruct the complex peripheral nerve anatomy as well as pinpoint regions of pathology. MRI can also be used to image signal changes in denervated muscle.

The severity or grade of a peripheral nerve injury is determined by the magnitude and duration of the applied forces of injury. Seddon defined three grades of nerve injury (neurapraxia, axonotmesis, and neurotmesis) based on the extent of injury to the three structural components of the peripheral nerve described above²¹.

Neurapraxia, the mildest grade of nerve injury, is characterised by a reduction or complete blockage of conduction across a segment of nerve. Axonal continuity is maintained and nerve conduction is preserved both proximal and distal to the lesion but not across the lesion. Neurapraxia can result from direct mechanical compression, ischaemia secondary to vascular compromise, metabolic derangements, or diseases or toxins causing demyelination of the nerve. Conduction is restored once either the metabolic derangement is corrected or remyelination occurs. Neurapraxic injuries are usually reversible and full recovery can occur within days to weeks.

Axonotmesis is a more severe grade of nerve injury and is characterised by interruption of the axons with preservation of the surrounding connective tissue “highway” which can support axonal regeneration. Distal Wallerian degeneration (axonal and myelin degeneration distal to the site of the injury) of the axons occurs over a period of several days, after which direct electrical stimulation of the disconnected distal nerve stump will not give rise to nerve conduction or muscle response. Recovery can occur through axonal regeneration due to the preservation of the connective tissue “highway” which consists of Schwann cells and their basal lamina. The Schwann cells proliferate and form longitudinal conduits through which axons regenerate. Axonotmetic injuries usually recover over a period of months. The timing and degree of recovery depends on several factors, including the extent of retrograde axonal loss, as well as the time to regenerate and re-innervate target muscles and/or sensory end organs. As a general rule, peripheral nerve fibres regenerate at a rate of approximately 1 mm/day or 1 inch/month. More proximal injuries, therefore, require a longer time for axonal regeneration to re-innervate their targets.

Neurotmesis is the severest grade of peripheral nerve injury. Neurotmetic injuries are characterised by disruption of the axon, myelin, and connective tissue “highway” components of the nerve. Recovery through regeneration cannot occur in this case. This grade of injury encompasses nerve lesions in which external continuity of the nerve is preserved but intraneural fibrosis occurs and blocks axonal regeneration. Neurotmetic injuries also include nerves whose continuity has been completely interrupted. Since the necessary “highways” for axonal regeneration are absent, surgery is required to remove any intervening roadblocks in the form of scar tissue as well as to reestablish continuity of the nerve.

On the basis of clinical symptoms and physical findings alone, it is often difficult to differentiate neurapraxic, axonotmetic, and neurotmetic grades of nerve injury, especially in the acute setting. Nerve conduction studies, both sensory and motor, are useful during and after the first week following an injury to distinguish neurapraxic from axonotmetic and neurotmetic injuries.

If the neurological examination shows signs of a nerve injury, the following diagnostic tests are recommended: EMG, nerve conduction studies and MRI. Reduced muscle activity in the EMG can indicate nerve injury. EMG is a diagnostic procedure to assess the health of muscles and the nerve cells that control them (motor neurons). A nerve conduction study uses surface electrodes (electrodes taped to the skin) to measure the speed and strength of signals travelling between two or more points. EMG results can reveal nerve dysfunction, muscle dysfunction or problems with nerve-to-muscle signal transmission. These diagnostic procedures are not different from those performed in non-haemophiliacs.

Treatment

Trauma is the most frequent cause of a peripheral nerve injury. Nerve injuries are caused by traction, compression, sharp laceration, and missiles (e.g. gunshot wounds). Traction injury is often associated with a fracture or dislocation. An understanding of the mechanism of injury is extremely helpful in determining the severity of the lesion and for guiding the clinical management.

Traumatic peripheral nerve injuries can be classified into open and closed injuries. Decision-making related to open injuries is relatively straightforward. Acute sharp lacerating injuries should repaired immediately, with the aim of primary end-to-end suture repair. However, not all transecting injuries lend themselves to primary repair. If the ends are ragged or contused, delayed repair is preferable to demarcate normal from abnormal neural tissue.

The decision-making process for the treatment of closed traumatic peripheral nerve injuries is more complex. The majority of closed traumatic injuries are due to stretching and/or compressive forces. An associated expanding haematoma producing a compartment syndrome may require emergency surgery to avoid irreversible nerve injury. In the majority of closed traumatic injuries, however, the nerves are not actually transected. Instead, a “lesion in continuity”, representing the damaged segment of nerve, may be produced which results in either a neurapraxic, axonotmetic, neurotmetic, or combination of these grades of injury.

If a nerve is injured but not divided into two, it must be assessed whether the nerve is healing. If the nerve is healing properly, which can take up to 2 years, there is a good chance that the patient will not need surgical repair. The patient will need regular check-ups to make sure his recovery continues to its maximum potential. Depending on the type and severity of the nerve injury, the patient may need pregabaline or gabapentin, paracetamol or COX-2 inhibitors, to relieve his pain. Pregabaline or gabapentin are often more useful for this purpose than paracetamol or COX2 inhibitors^22–24. The patient may also need physical therapy and orthopaedic appliances to prevent stiffness.

Non-operative treatment

The following treatments can help restore function to the affected muscles: braces or splints, electrical stimulation, pulsed ultrasound, and physical and occupational therapy. Braces or splints help to hold the affected limb, fingers, hand or foot in the proper position to improve muscle function.

According to Ryan et al. a considerable proportion of paediatric peroneal palsies arise as iatrogenic complications of casting, footboards, and intraoperative positioning²⁵. They described a patient who developed bilateral peroneal palsies after cardiac transplantation. Factors that predisposed this patient to the development of a neuropathy included haemophilia and prolonged immobilisation, with focal involvement of the peroneal nerves being likely related to the use of ankle-foot orthoses. Peroneal nerve palsies are a significant, but potentially avoidable, iatrogenic complication of lower-extremity orthoses.

Electrical stimulators can activate the involved muscle served by an injured nerve while the nerve regrows. However, this treatment may not be effective for everyone. Physical and occupational therapy involves specific movements or exercises to keep the affected muscles and joints active. Such therapy can prevent stiffness and help to restore function and feeling.

In their series, Saraf et al. aimed to highlight the role of pulsed ultrasound in the management of 134 patients with bleeding disorders and nerve palsies (n=20)¹. In addition to usual conservative treatment, 10–15 sessions of pulsed ultrasound electrotherapy were used in all the cases of nerve palsy. Twelve cases improved completely within 6 months whereas another four recovered around 10 months. The remaining four showed only partial recovery even when followed up to 12 months. Recognition of nerve complications in haemophiliacs is important to avoid long-term disability. Conservative therapy starting early is commonly associated with favourable results. Pulsed ultrasound therapy at the site of haematoma may help in the early regression of haematoma and neural recovery.

Initial treatment of most compressive neuropathies is non-operative, consisting of rest, avoidance of joint movement, and, when necessary, temporary immobilisation of the involved joint²⁶. If symptoms persist, particularly when accompanied by muscle weakness, surgery is usually indicated. In the case of compression or stretching, it is often not possible to determine the grade of the injury immediately. A partial nerve injury associated with muscle denervation usually indicates an axonotmetic grade of injury. Patients with such injuries should be followed with serial clinical and electrodiagnostic examinations to document recovery and confirm the diagnosis. These patients do not require immediate surgical intervention. An enlarging haematoma can convert a partial nerve injury into a complete injury. Complete nerve injuries produce severe muscle denervation and may represent either an axonotmetic or neurotmetic grade of injury. It is critical to distinguish between these two grades of injury over time, since the latter is dependent on surgical repair for recovery. Patients are, therefore, followed closely for several months, looking for clinical and electrodiagnostic evidence of nerve regeneration and muscle reinnervation. Muscles should be re-innervated within 2 years following a traumatic nerve injury if recovery of useful motor function is to occur.

Surgical treatment

After 2 years, denervated muscles undergo irreversible atrophy and replacement by fat. It is, therefore, necessary to time surgical exploration so that a successful nerve repair results in muscle re-innervation within 2 years of the injury. A useful rule of thumb is to follow a patient for 3 to 4 months to allow any element of neurapraxia to resolve as well as to enable axonal regeneration to occur beyond the point of injury. If there is no clinical or electrodiagnostic evidence of muscle re-innervation, then surgical exploration using intraoperative electrophysiological monitoring should be performed. Another approach in the management of traumatic peripheral nerve injuries is to operate “early” (i.e. as soon as medically feasible). The rationale for this approach includes less scarring and, therefore, easier dissection of peripheral nerve elements and intraoperative evaluation of anatomical and electrophysiological continuity. However, it remains controversial whether earlier surgical repair leads to a better recovery of peripheral nerve function.

If one of the peripheral nerves fails to heal properly, the surgeon can use EMG testing in the operating room to assess whether scarred nerves are recovering. Doing an EMG test directly on the nerve is more accurate and reliable than doing the test over the skin.

Sometimes a section of a nerve is cut completely or damaged beyond repair. The surgeon can remove the damaged section and reconnect healthy nerve ends (nerve repair) or implant a piece of nerve from another part of the body (nerve graft). Those procedures can help the patient’s nerves to regrow. Sometimes the surgeon can borrow another working nerve to make an injured nerve work (nerve transfer).

If the patient has a particularly severe nerve injury or one that has been untreated for a long time, surgery is indicated to restore function to critical muscles by transferring tendons from one muscle to another.

Novel strategies in the repair of injured peripheral nerves with ethyl-cyanoacrylate adhesive showed functional recovery comparable to that of conventional epineural sutures²⁷.

There is no easily available formula for the successful treatment of peripheral nerve injuries. Incomplete injuries are more frequent²¹. After complete axonal transection, the neuron undergoes a number of degenerative processes, followed by attempts at regeneration. A distal growth cone seeks out connections with the degenerated distal fibre. The current surgical standard is epineural repair with nylon sutures. To span gaps that primary repair cannot bridge without excessive tension, nerve-cable interfascicular auto-grafts are employed. Unfortunately, the results of nerve repair to date have been modest, with only 50% of patients regaining useful function. There is much ongoing research regarding pharmacological agents, immune system modulators, enhancing factors, and entubulation chambers. Clinically applicable developments from these investigations will continue to improve the results of treatment of nerve injuries.

Conclusions

The most frequent nerve injuries in haemophilia are femoral nerve neuropathy due to iliopsoas haematoma, ulnar nerve neuropathy at the elbow due to cubitus valgus in cases with severe deformity of the elbow associated with advanced haemophilic arthropathy, carpal tunnel syndrome and peroneal nerve palsy as a complication of TKA. Pain around the hip joint, femoral neuropathy and hip flexion contracture in a patient with haemophilia should alert the physician to the possibility of an iliopsoas haematoma. Early and effective factor replacement therapy is essential in the prevention of the complications. I believe that the presence of an inhibitor may not be, per se, a risk factor for peripheral nerve injuries in haemophilia. However, if adequate haemostasis is not achieved, the patient could be prone to abnormal muscle bleeding and this could increase the risk of nerve injury in patients with inhibitors. The best way to manage nerve injuries in haemophilia is prevention of muscle haematomas and haemophilic arthropathy.

The initial treatment of most compressive neuropathies is non-surgical, conservative management (braces-splints-orthoses, electrical stimulation, pulsed ultrasound, and physical and occupational therapy). If symptoms persist, surgery is usually indicated. In the case of compression or stretching, it if often not possible to determine the grade of the injury immediately. A partial nerve injury associated with muscle denervation usually indicates an axonotmetic grade of injury. Patients with such injuries should be followed with serial clinical and electrodiagnostic examinations to document recovery and confirm the diagnosis. These patients do not require immediate surgical intervention. An enlarging haematoma can convert a partial nerve injury into a complete injury. Complete nerve injuries produce severe muscle denervation and may be either an axonotmetic or neurotmetic grade of injury. It is critical to distinguish between these two grades of injury over time, since the latter requires surgical repair for recovery to occur. Muscles should be re-innervated within 2 years following a traumatic nerve injury if recovery of useful motor function is to occur.