Introduction

Malignant peripheral nerve sheath tumors (MPNSTs) originate from the cells that enclose the axons of peripheral nerves. The majority of MPNSTs in the dog are presumed to arise from Schwann cells; however, they often have a remarkably varied histological appearance with mixed cell populations.¹ This has led to a varied array of nomenclature describing MPNSTs, but the umbrella term of MPNST has been generally accepted due to their similar biological behavior and progression.²

Although MPNSTs are often locally invasive within the affected nerve and have malignant histological characteristics, pulmonary metastasis is rare.^3–5 In dogs, they most frequently arise from the brachial plexus and cervical and cranial thoracic spinal nerve roots (C6–T2), but have also been reported in other spinal nerve roots and distal peripheral nerves.^3,4,6–9 Clinical signs reflect the location of the neoplasm and commonly have an insidious onset including progressive lameness, muscle atrophy, and lower motor neuron and proprioceptive deficits.^4,5,9

Successful surgical treatment is dependent on tumor location, with local resection, amputation, or laminectomy having been described.^3–5,10 MPNST often carries a poor prognosis due to the difficulty in controlling local disease; hence, tumor location has been shown to have a significant effect on survival time. MPNSTs located in distal peripheral nerves carry a more favorable prognosis than those affecting the brachial plexus or nerves near the vertebral canal.⁴

The dog described here was diagnosed with MPNST affecting the deep branch of the radial nerve, distal to the elbow. Nerve decompression through epineurotomy improved clinical signs for 3 mo; however, thoracic limb amputation was ultimately performed due to recurrence of clinical signs.

Case Report

A 4 yr 3 mo old male neutered Labrador retriever was referred for investigation of chronic thoracic limb lameness. The patient initially presented to the primary veterinarian 6 mo prior to referral with a mild intermittent right thoracic limb lameness, which was treated conservatively with exercise restriction and meloxicam. Non-weight bearing lameness recurred intermittently, and pain could be elicited on extension of the right shoulder and palpation of the right humeral condyle. Mild atrophy of the right shoulder musculature was documented at this time, but survey radiographs of the right shoulder and elbow did not show any abnormality. Further investigations were performed 3 mo after initial presentation due to worsening lameness and pain. These included MRI of the cervical spine, brachial plexus, and elbow, motor nerve conduction velocity, electromyography, and arthrocentesis of the elbow; however, these were normal. Treatment with gabapentin^a (10 mg/kg PO TID), tramadol^b (3 mg/kg PO TID), and meloxicam^c (0.1 mg/kg PO SID) was instigated, but meloxicam^c was discontinued 2 days later due to an episode of hemorrhagic diarrhea. The dog remained persistently lame with intermittent episodes of acute vocalization or pain for which the owner administered tramadol^b (3mg/kg PO TID) and diazepam^d (0.5mg/kg PO PRN). Three months later, the patient was referred due to progressive deterioration in the lameness.

On referral to the orthopaedic service, the patient weighed 32.7 kg with a body condition score of 3/5. General physical examination was within normal limits, and mentation was normal. On orthopaedic examination, the patient was non-weight-bearing lame on the right thoracic limb with severe atrophy of the supra and infraspinatus muscles. No crepitus or decreased range of motion was detected in the joints, although marked pain was elicited on flexion and extension of the right elbow and on deep palpation of the soft tissues overlaying the cranial aspect of the proximal radius (region of extensor carpi radialis muscle) and cranial elbow. The rest of the orthopaedic examination was unremarkable. Neurological examination revealed reduced withdrawal reflex, but sensation was normal. It was not possible to interpret hopping reactions due to non-weight-bearing lameness, and paw positioning was not possible to assess, as he would not place the paw on the ground. The rest of the neurological examination was normal.

The patient was hospitalised and a 20G cannula was placed in the left cephalic vein. Complete blood cell count and serum biochemical profile were unremarkable. The patient was sedated with medetomidine^e (10 μg/kg IV) and methadone^f (0.2μg/kg IV), and computed tomography (CT) of the thorax and right thoracic limb from shoulder to mid-antebrachium was performed (Philips Brilliance CTMx8000 IDT, 16 slice scanner). Thoracic CT was normal, and the limb CT showed marked muscle atrophy and no osseous abnormalities. CT was repeated following intravenous injection of 2 ml/kg non-ionized iodine contrast (Omnipaque^g 300 mg/ml), which revealed a 1 cm area of contrast enhancement in the region of the radial nerve just distal to the elbow. Ultrasonography of the region revealed a 4 mm focal enlargement of a branch of the radial nerve adjacent to the lateral aspect of the proximal radius (Figure 1). The patient was discharged whilst diagnostics were reviewed and medical management with tramadol^b (3mg/kg PO BID), gabapentin^a (10 mg/kg PO BID), and diazepam^d (0.5 mg/kg PO PRN) was continued.

View Figure

• Download as Powerpoint
• Download Figure

Given the high index of suspicion for MPNST, limb amputation was recommended; however, the owner declined. Permission was granted for exploratory surgery to biopsy the lesion in order to obtain a definitive diagnosis. The patient was premedicated with intravenous acepromazine^h (0.03 mg/kg) and methadone^f (0.2 mg/kg). General anaesthesia was induced with intravenous propofolⁱ (5 mg/kg) and maintained with inhalation isoflurane^j and oxygen. Meloxicam^c (0.1 mg/kg IV) was administered before surgery.

The right thoracic limb was aseptically prepared, and a craniolateral approach to the radial head was performed. The extensor carpi radialis muscle was retracted medially, and a mass could be palpated beneath the supinator muscle. These muscles were then retracted laterally to allow the mass to be accessed within the muscular branch of the radial nerve to the extensors (Figure 2). The epineurum was opened and several biopsies were taken with metzenbaum scissors. A local splash block was performed using ropivacaine^k prior to routine closure. There was a significant improvement in the dog's lameness and comfort immediately post-operatively. The patient was hospitalized for 2 days and analgesia provided with methadone^f (0.2 mg/kg IV q4 hours). The dog was discharged with instructions of exercise restriction for 6 wk and continuation of the same medical therapy as previously (gabapentin^a, tramadol^b, and diazepam^d).

View Figure

• Download as Powerpoint
• Download Figure

Histopathological analysis of the sample was consistent with a diagnosis of MPNST. The tissue consisted of a non-encapsulated infiltrate of a moderately cellular neoplastic spindeloid cell population.

Following histopathology results, amputation of the right thoracic limb was recommended as soon as possible to maximize chances of obtaining clean margins and reducing the risk of metastasis and proximal infiltration of the radial nerve. The dog was assessed 3 wk post-operatively to be mildly lame on the right thoracic limb, and the previously documented muscle atrophy had improved. Limb amputation was again declined by the owner due to the marked improvement in the patient's limb function. The analgesic treatment was progressively reduced and the owner allowed periods of swimming to increase mobility and muscle mass in the affected limb.

The owner reported a gradual reduction in use of the right thoracic limb from approximately 3 mo after the biopsy procedure, which progressively deteriorated despite increasing the frequency of gabapentin^a to three times daily. On representation to the referral clinic 4.5 mo after surgery, there was a non-weight-bearing lameness, with the dog occasionally toe-touching, and pain was elicited on palpation of the proximal radius. Atrophy of the shoulder muscles was once again evident. An ultrasound of the affected radial nerve revealed enlargement of the mass to 8 mm with thickening of the nerve proximally indicating progressive spread of the neoplasm. Limb amputation was recommended in the face of disease progression, but due to the improvement following epineurotomy at the time of biopsy, the owner declined limb amputation and elected for repeat exploratory surgery. The risk of failure to improve clinical signs and the likely detrimental consequences of delaying excision were explained in detail.

The patient was anaesthetised and prepared following the same protocol as previously. A minimal craniolateral approach to the deep branch of the radial nerve was performed. The normal anatomy over the tumor site was distorted due to fibrous scarring. The mass had progressed locally and the nerve proximal to it was grossly abnormal, extending proximal to the elbow at the level of the deep branch of the radial nerve to the extensor carpi radialis muscle. The thickened nerve sheath was incised proximally beyond the interface with grossly normal tissue. Closure was routine.

Following epineurotomy there was improvement in lameness and pain, but to a lesser extent than previously. These signs once again deteriorated 1 mo post-operatively, and the owner elected for limb amputation at that time. Due to financial constraints, this was performed by the referring veterinarian without complication. Histopathological analysis of the limb confirmed malignant peripheral nerve sheath tumor of the radial nerve with clean margins of excision. Three weeks after surgery, the owners reported that the patient had recovered well from surgery and had a good quality of life. Annual re-examination has been recommended, with survey thoracic CT to monitor for signs of pulmonary metastasis.

Eighteen months after amputation, the patient was clinically well with no signs of recurrence or metastasis.

Discussion

Epineurotomy as a palliative treatment for limb MPNST has not been previously described in dogs. Since this neoplasia tends to be locally invasive, of slow progression, and with low metastatic rate, this procedure may provide some palliation of clinical signs in cases in which the tumor is located in the distal portion of a peripheral nerve and amputation delay will not necessarily compromise the likelihood of achieving complete margins, or in those cases in which amputation is simply not possible.

The prognosis for MPNST in dogs depends, in part, on tumor location and whether or not the tumor is amenable to surgery, with very limited information on the results of adjunctive therapy. Studies describing cases treated surgically reported a poor prognosis and high rates of recurrence, which ultimately lead to failure due to the associated neurological signs.^3–5 Several studies have shown variation in survival time with location of the lesion, with tumors either affecting the nerve roots or invading the intervertebral foramina carrying a worse prognosis than those affecting the brachial plexus.^4,5 Brehm and others (1995) further described that tumors distal to the brachial or lumbosacral plexus had a far more favorable prognosis than MPNST located more proximally; however, due to their infrequent occurrence, a mean survival time is not known. Of the five dogs with distal MPNST that underwent local resection, two were lost to follow up and the remaining three dogs died of unrelated disease at 3 mo, 7 mo, and 7 yr post-operatively.⁴

Nerve compression causes endoneurial edema, constriction of epineurial blood flow, demyelination, and fibrosis with increasing pressure, which result in neuropathic pain and nerve dysfunction.¹² This would be consistent with the insidious onset of clinical signs seen with MPNST as the mass expands within the constrictive epineurium. Atrophy of the musculature of the proximal limb in this case is likely to be a result of disuse, which subsequently improved as limb function returned. The significant clinical improvement following epineurotomy was likely a result of decompression of the radial nerve caused by the growing lesion within the epineurium. The degree of compression would depend on a number of factors, including location of the lesion and rate of tumor growth. It is likely that a degree of neuropathy would develop as a result of the compression of the nerve as well as due to sensory nerve involvement. A similar presentation is seen in progression of human patients with carpal tunnel syndrome caused by compression of the median nerve at the wrist. Initial symptoms such as intermittent paresthesia and deficits in sensation early in the course of disease result from changes in intraneural microcirculation due to edema.¹² As the condition progresses, morphological changes such as segmental demyelination and Wallerian degeneration occur, resulting in constant pain and permanent sensory dysfunction.¹² At the time of the second surgery in this dog, the mass had propagated proximally, and there was marked fibrosis in the region. Progressive permanent structural degeneration of the nerve may explain the reduced success of repeat epineurotomy in this case.

The use of CT and ultrasound has been previously described in the evaluation of MPNSTs in humans and dogs; however, MRI has been shown to be more sensitive for detecting brachial plexus masses in humans.^13–17 Kraft and others described the importance of positioning in dorsal recumbency with symmetrically positioned forelimbs and upright, non-oblique body position allowing for comparison with the contralateral unaffected limb.⁸ The lesion was not detected on MRI 3 mo prior to presentation at the referral institution, as the image only extended to the level of the elbow. In this case, ultrasonography was vital in identifying the extent of the lesion as well as documenting its progression over time.

MPNSTs affecting the brachial plexus or peripheral nerves can be treated with local excision or amputation.^3–6,10 Hemilaminectomy is indicated in cases with infiltration of the nerve root and may be performed in conjunction with amputation. Adjunctive full course fractionated radiation therapy has been shown to be highly effective in the treatment of postoperative residual microscopic disease for a variety of neoplasias that, like MPNST, are included in the soft tissue sarcoma group; however, only anecdotal reports specifically about MPNSTs exist at this time.¹¹

In the present case, local resection of the radial nerve at the level of the elbow would have resulted in denervation of the carpal extensor muscles and would have necessitated additional procedures such as pancarpal arthrodesis in order to maintain any limb function. This degree of morbidity to the patient was deemed unacceptable; therefore, limb amputation early in the course of treatment was recommended to maximize the chances of complete excision, reduce risk of recurrence, and control clinical signs. The owners' initial refusal to permit limb amputation meant that options for “limb-sparing” palliative treatments were explored. Medical therapy with biopsy and epineurotomy performed through a minimal approach to the site in the region of the lesion seemed a reasonable option for diagnosis and possible palliation of clinical signs. Postoperative improvement in pain and lameness was more marked than initially expected. Medical therapy alone is usually not successful in the long term and, in this patient, had already failed prior to epineurotomy. Although the mechanism is not well understood, hypofractionated radiation therapy has been described as effective for palliation of somatic and neuropathic pain secondary to canine bone cancer and for somatic pain associated to non-resectable limb soft tissue sarcomas.^18,19 Its use could have been considered in this case, although it would have required the dog to have multiple general anaesthetics, and specific information about success rate of neuropathic pain control secondary to MPNST is not available in the literature.

This case demonstrates that, as expected, delayed attempts for complete excision of the MPNST lead to progression and subsequent proximal infiltration of the nerve. If left untreated, the neoplasm would spread proximally with increased risk of recurrence, even following amputation, and an opportunity for long-term control might be missed⁴. Significant gross changes occurred during the 4 mo between surgeries when compared with the preceding 6 mo since clinical signs began, growing from 4 mm at the first surgery to 8 mm at the second. The apparent increase in growth rate may be due to increased tumor growth, but may also be due to scarring and fibrosis following the initial epineurotomy. Little published information is available regarding long-term behavior of appendicular MPNST without radical surgical therapy. However, in a case series of 10 dogs with trigeminal MPNSTs, the seven untreated dogs' survival time was up to 21 mo from onset of clinical signs.²⁰ The limiting factor for survival time in these dogs was poor quality of life due to progression of clinical signs with increased tumor growth. In the case presented here, a second palliative epineurotomy led to only limited improvement for 1 mo. Amputation was ultimately performed as initially recommended and resulted in the dog being disease-free for the follow-up period of 18 mo.

Conclusion

In cases of MPNSTs when recommended radical surgery (limb amputation or complete local resection) cannot be performed, epineurotomy may be considered as a surgical option to palliate clinical signs. This consideration must be carefully balanced with morbidity to the patient and their anticipated quality of life post-operatively as well as the likelihood of proximal tumor progression and potential increased risk of recurrence should the lesion be subsequently excised. This case would indicate that repeat epineurotomy may achieve some degree of clinical improvement; however, this may not be to the same extent or duration as that seen following the initial intervention.