Key points

Introduction

Neuromuscular complications related to cancer are common. Cancer can directly affect the peripheral nervous system at any level via numerous mechanisms, including direct nerve compression or infiltration, hematogenous or lymphatic spread, meningeal dissemination, or perineural spread. Paraneoplastic syndromes often manifest with neuromuscular dysfunction, as can cancer-associated medical complications, such as infections, weight loss, or malnutrition. Acquired neuropathies can result from effects of cancer treatment itself, be it surgery, chemotherapy, radiation therapy, hematopoietic stem cell transplantation, or immunologic therapy. Patients may also have pre-existing neurologic conditions, such as diabetic or hereditary neuropathies, that can be exacerbated by cancer or its related treatments. Often, a combination of processes can be present.

Electrodiagnostic studies, including nerve conduction studies (NCS) and needle electromyography (EMG), are invaluable tools for assessing neuromuscular function in patients with cancer. Electrodiagnosis can confirm a suspected neuropathic or myopathic process as well as rule out other possibilities. It can detect subclinical neuropathies, which can inform clinical decision making regarding use of neurotoxic chemotherapeutic agents. They can help with localizing lesions and determining pathophysiology, chronicity, and severity, which in turn can aid the cancer physiatrist in determining prognosis for recovery and the utility of future rehabilitation interventions. Finally, the information obtained with electrodiagnostic testing can help guide the oncology team with regards to surgery, chemotherapy, or radiation therapy planning.

Electrodiagnosis should be thought of as an extension of the history and physical examination, with the expected clinical and NCS/EMG findings dependent on the location, distribution, and pathophysiology of the neurologic lesion. Any and all levels of the peripheral nervous system can be affected by cancer and its treatments, including spinal roots, brachial or lumbosacral plexus, peripheral axons and/or myelin sheaths, the neuromuscular junction, and muscle fibers. Because of the variety of mechanisms of injury and wide scope of clinical presentation, the true incidence and prevalence of neuromuscular disorders in patients with cancer are unknown. However, it is estimated that approximately one-third of adult patients with chronic cancer pain, across all tumor types and stages, are thought to have cancer-related neuropathic pain.

Introduction

Neuromuscular complications related to cancer are common. Cancer can directly affect the peripheral nervous system at any level via numerous mechanisms, including direct nerve compression or infiltration, hematogenous or lymphatic spread, meningeal dissemination, or perineural spread. Paraneoplastic syndromes often manifest with neuromuscular dysfunction, as can cancer-associated medical complications, such as infections, weight loss, or malnutrition. Acquired neuropathies can result from effects of cancer treatment itself, be it surgery, chemotherapy, radiation therapy, hematopoietic stem cell transplantation, or immunologic therapy. Patients may also have pre-existing neurologic conditions, such as diabetic or hereditary neuropathies, that can be exacerbated by cancer or its related treatments. Often, a combination of processes can be present.

Electrodiagnostic studies, including nerve conduction studies (NCS) and needle electromyography (EMG), are invaluable tools for assessing neuromuscular function in patients with cancer. Electrodiagnosis can confirm a suspected neuropathic or myopathic process as well as rule out other possibilities. It can detect subclinical neuropathies, which can inform clinical decision making regarding use of neurotoxic chemotherapeutic agents. They can help with localizing lesions and determining pathophysiology, chronicity, and severity, which in turn can aid the cancer physiatrist in determining prognosis for recovery and the utility of future rehabilitation interventions. Finally, the information obtained with electrodiagnostic testing can help guide the oncology team with regards to surgery, chemotherapy, or radiation therapy planning.

Electrodiagnosis should be thought of as an extension of the history and physical examination, with the expected clinical and NCS/EMG findings dependent on the location, distribution, and pathophysiology of the neurologic lesion. Any and all levels of the peripheral nervous system can be affected by cancer and its treatments, including spinal roots, brachial or lumbosacral plexus, peripheral axons and/or myelin sheaths, the neuromuscular junction, and muscle fibers. Because of the variety of mechanisms of injury and wide scope of clinical presentation, the true incidence and prevalence of neuromuscular disorders in patients with cancer are unknown. However, it is estimated that approximately one-third of adult patients with chronic cancer pain, across all tumor types and stages, are thought to have cancer-related neuropathic pain.

Radiculopathy

After disc disease and spinal stenosis, tumors involving the spine and spinal cord are the most common causes of radiculopathy. All tumor types can metastasize to the spine, although the most common primary malignancies that do so include breast, lung, prostate, colon, thyroid, and kidney. Common primary malignant spinal tumors include multiple myeloma, plasmacytoma, and Ewing and osteogenic sarcoma. Single- or multilevel radiculopathies due to malignancy can result from primary or epidural metastatic tumor extension into the neural foramina. Leptomeningeal disease is due to metastatic involvement of the leptomeninges from infiltrating cancer cells, and involvement of the cauda equina can be thought of as a lumbosacral polyradiculopathy. The most common primary cancers associated with leptomeningeal disease are breast, lung, gastric, melanoma, lymphomas, and leukemias. Of the leukemias, leptomeningeal disease is most commonly seen in acute lymphocytic leukemia.

Patients can present with an asymmetric array of symptoms resulting from radicular or polyradicular involvement, including focal and radicular pain, areflexia, paresthesias, and lower motor neuron weakness. In leptomeningeal disease, there may be associated findings of nuchal rigidity as well as upper motor neuron signs, especially if there is concomitant brain involvement. Cranial nerves can be involved as well, with the oculomotor, facial, and auditory nerves most commonly affected.

In radiculopathies, sensory responses should be normal on NCS, because the location of involvement is proximal to the dorsal root ganglion, thereby making the segment of sensory nerve fibers tested metabolically and histologically intact. Motor responses within the affected myotomes may be normal or reduced in amplitude, depending on severity. Needle EMG is the most sensitive electrodiagnostic test for evaluation of a radiculopathy. One should record neuropathic abnormalities in at least 2 muscles innervated by different peripheral nerves but sharing the same root innervation, including increased insertional activity, fibrillation potentials, reduced recruitment, and large, polyphasic motor unit potentials (MUPs). Because paraspinal muscles are innervated by the dorsal primary rami, branching directly off of the nerve root, abnormal neuropathic EMG findings noted in the paraspinals further support the diagnosis of radiculopathy. Electrodiagnostic studies in leptomeningeal disease can sometimes be consistent with a polyradiculopathy, with preserved SNAPs and abnormal paraspinal needle EMG findings. Absent F waves or prolonged F-wave latencies on NCS are thought to be an early indicator of nerve root involvement, but are not specific for either radiculopathy or leptomeningeal disease.

Plexopathy

Brachial plexopathies from neoplasms are usually the result of metastatic disease, with breast and lung being the most common primary sources. Symptoms include pain, paresthesias, numbness, and weakness in the distribution of plexus involvement. Metastases can involve any portion of the brachial plexus, but usually involve the lower trunk preferentially, because of its proximity to axillary lymph nodes and the superior sulcus of the lung. Neoplastic lumbosacral plexopathies can also stem from metastatic disease, but are much more likely to be caused by direct extension of local tumor or perineural spread. Common tumors involved in lumbosacral plexus injury include colon, gynecologic tumors, lymphomas, and sarcomas.

The primary differential diagnostic concern in a patient with cancer with a plexus injury is distinguishing between a neoplastic and radiation-induced cause. Occasionally, the 2 conditions can coexist. Classically, radiation-induced plexopathy is delayed in onset, pain is less common than in neoplastic plexopathy, and symptoms of weakness and paresthesias are usually progressive. There is also more likely to be associated lymphedema in the involved limb. It has been reported that neoplastic brachial plexopathy tends to preferentially affect the lower trunk, and radiation plexopathy tends to preferentially affect the upper portion of the plexus; however, further studies suggest that plexus involvement may be more diffuse and with more overlap in both causes than previously suspected.

The distribution of motor and sensory nerve conduction abnormalities is important in the localization of both brachial and lumbosacral plexopathies. For instance, upper extremity NCS in a lower trunk brachial plexopathy will demonstrate a characteristic pattern of normal median sensory nerve action potential (SNAP) amplitudes, reduced ulnar SNAP amplitudes, and reduced median and ulnar compound muscle action potential (CMAP) amplitudes. Abnormalities will also be noted in the medial antebrachial cutaneous SNAP. Findings on needle EMG will demonstrate fibrillation potentials, reduced recruitment, and large, polyphasic MUPs within the distribution of involvement. Needle EMG of the paraspinal muscles is normal in a pure plexopathy; however, one often sees both root and plexus involvement in a given patient, depending on the extent of disease. Documenting asymmetric findings on NCS and needle EMG can sometimes help distinguish a newer onset radiculopathy or plexopathy in the setting of an underlying chemotherapy-induced polyneuropathy.

Approximately 50% of all patients with cancer will undergo radiation therapy at some point during the course of their disease, and radiation therapy is involved in approximately one-quarter of all cancer cures. As patients are living longer following cancer treatments, physicians are becoming more aware of late neuromuscular complications of therapy, especially radiation therapy. Side effects are essentially related to the dose of radiation and the volume of normal tissue that receives radiation.

The pathognomonic EMG finding of radiation-induced neuropathic damage is the myokymic discharges, which are clusters of MUPs firing spontaneously with regular interburst and intraburst frequencies. The absence of myokymic discharges however does not exclude radiation damage, and although the presence of myokymic discharges and fasciculation potentials confirms a radiation-induced contribution to plexus injury, it does not exclude the possibility of tumor involvement. Even in the setting of classic EMG findings, follow-up imaging of the brachial plexus with MRI is warranted to exclude a concomitant compressive or infiltrating lesion, which could be due to local recurrence or new metastases. In a patient with a history of prior radiation therapy to the axillary or supraclavicular lymph nodes, secondary radiation-induced neoplasms such as sarcomas should also be considered.

Mononeuropathy

Focal mononeuropathies directly related to cancer most often result from the external compression or invasion from tumor, such as an isolated radial neuropathy caused by a primary osteogenic sarcoma or a bone metastasis involving the spiral groove of the humerus. Malignant nerve sheath tumors arising from plexiform neurofibromas can also result in focal mononeuropathies. The presence of secondary sarcomas compressing or infiltrating nervous system structures, resulting from previous radiation therapy, needs to be excluded. NCS and needle EMG should correspond to clinical abnormalities, limited to the distribution of the individual nerve, involving both sensory and motor fibers, depending on the composition of the particular nerve involved.

Although uncommon, damage to the peripheral nervous system during the perioperative period can occur in the surgical patient with cancer. Because the nature of these procedures is likely to be more complex than in nononcologic surgeries, it is postulated that the likelihood of neuromuscular complications is greater in oncologic surgeries. There are no studies, however, comparing the incidence of unintentional nerve injury in the cancer surgery population to that in the general population. In addition, peripheral nerves are sometimes intentionally sacrificed in the cancer surgery patient in order to obtain local disease control, especially the spinal accessory nerve to the trapezius in radical or modified neck dissection for head and neck malignancies or in limb-salvage surgery for extremity sarcomas. The pattern and extent of neurologic involvement following surgery depends on the location of the tumor, patient positioning during surgery, and the patient’s overall preoperative status and propensity to nerve injury.

Perioperative neurapraxic injuries, resulting from either compression or traction of peripheral nerves, are well-recognized phenomena. It is thought that these injuries result from the patient’s position during anesthesia and surgery or during the immediate postsurgical recovery period. Common sites of injury and associated surgical procedures include brachial plexus injury during thoracotomy or mastectomy, given the abducted position of the involved upper extremity. Abduction of the upper extremity greater than 90° during surgery can cause the humeral head to sublux inferiorly, resulting in compression of the lower part and traction of the upper part of the brachial plexus. Patients upon awakening report varying degrees of pain, weakness, and numbness in both the upper and the lower trunk distribution. Complete, spontaneous recovery within weeks is common, even in cases of severe plegia. Ulnar neuropathies at the elbow, resulting from arm boards used to secure intravenous lines, and radial neuropathies at the spiral groove, resulting from prolonged time in the lateral decubitus position, are also noted following thoracic surgery. Findings of focal slowing, temporal dispersion, or conduction block across the level of injury can be demonstrated on motor NCS.

Compression of the femoral nerve or lumbar plexus can result from traction during pelvic surgery. Patients undergoing hip arthroplasty or acetabular reconstruction are especially susceptible to injury of the peroneal division of the sciatic nerve. Injuries to the superior gluteal, obturator, and femoral nerves have also been reported. A patient with cancer having undergone a significant amount of weight loss may be more susceptible to a perioperative peroneal neuropathy at the fibular head, resulting from positioning following a prolonged surgery and postoperative recovery period. Finally, delayed postoperative hemorrhages and hematomas should be excluded in all patients who develop new neuropathic symptoms 24 to 48 hours after surgery.

Rapid weight loss is a common symptom of malignancy, and thus, there is a higher risk of focal compression neuropathy, especially the peroneal nerve at the level of the fibular head. The peroneal nerve in that location is no longer protected by soft tissue and is more easily compressed against bony structures. A history of habitual leg crossing is sometimes elicited. Focal slowing, temporal dispersion, and conduction block across the fibular head on peroneal motor NCS are characteristic findings. There may be additional predisposition to injury given exposure to neurotoxic chemotherapy.