USMLE – Investigation of Neurological Disease
TESTS OF FUNCTION (CLINICAL NEUROPHYSIOLOGY)
In the investigation of neurological disease, tests of function have a somewhat more restricted application than tests of structure (i.e. imaging). Nevertheless, recording of electrical activity over the brain and assessment of nerve and muscle function are essential in certain conditions. The major tests are electroencephalography (EEG), evoked potentials (EPs), and nerve conduction studies/electromyography (NCS/EMG).
Electrical activity arising in the cerebral cortex can be detected using electrodes placed on the scalp, although this is estimated to detect only 0.1-1% of the brain’s electrical activity at anyone time. An army of electrodes provides spatial information. Rhythmical wave-forms can be detected and are distinguished by their frequency. When the eyes are shut, the most obvious frequency over the occipital cortex is 7-13/s; this is known as alpha rhythm and disappears when the eyes are opened. Other frequency bands seen over different parts of the brain in different circumstances are beta (faster than 13/s), theta (4-6/s), and delta (slower than 4/s). Lower frequencies predominate in the very young and during sleep.
Various diseases result in abnormalities of the EEG. These may be continuous or episodic, focal or diffuse. Examples of continuous abnormalities include a global increase in fast frequencies (beta) seen with sedating drugs (e.g. benzodiazepines), or marked slowing seen over a structural lesion such as a tumor or an infarct. With the advent of modem neuro-imaging, EEG has lost its use in localizing lesions (except in the management of epilepsy), but it is still useful in the management of patients who have disturbance of consciousness or disorders of sleep, in the diagnosis of cerebral diseases such as encephalitis and in certain dementias (e.g. Creutzfeldt-Jakob disease).
The most important use of EEG is in the management of epilepsy. It must be stressed, however, that only in rare circumstances will an EEG provide unequivocal evidence of epilepsy, and it is therefore not useful as a diagnostic test for the presence of epilepsy. Its use is predominantly to distinguish the type of epilepsy present and whether there is an epileptic focus, particularly if surgery for epilepsy is contemplated.
During an epileptic seizure, high-voltage ‘transients’ can be recorded. These may be generalized, as in the 3 cycle/s ‘spike and wave’ of childhood absence epilepsy (petit mal), or more focal in partial epilepsies. However, it is unusual to record a seizure itself, except in the case of childhood absence epilepsy. Nevertheless, it is often possible to detect ‘epileptiform’ abnormalities in between seizures in the form of ‘spikes’ and ‘sharp waves’ which lend support to a clinical diagnosis. The likelihood of detecting these abnormalities is enhanced by hyperventilation, photic flicker, sleep and some drugs. Note that even so, some 50% of patients with proven epilepsy will have a normal ‘routine’ EEG, and conversely, the presence of features often seen in association with epilepsy does not, of itself, make a diagnosis (although the false positive rate for clear-cut epileptiform features is < 1/1000).
It is possible to enhance the information provided by a variety of means. For example, the usual 30 minute recording session can be lengthened to 24 hours by the use of a light-weight tape recorder. The addition of video information to the EEG allows comparison of behavior with cerebral activity. In special circumstances, electrodes can be surgically positioned, e.g. through the foramen ovale, to record from the inferior temporal surface.
If a stimulus is provided-for example, to the eye-it would normally be impossible to detect the small EEG response evoked over the occipital cortex as the signal would be lost in background noise. However, if the EEG data from 100-1000 repeated stimuli are averaged electronically, this noise is removed and an evoked potential recorded whose latency and amplitude can be measured.
Evoked potentials can be measured following visual, auditory or somatosensory stimuli if electrodes are appropriately positioned, though visual evoked potentials are by far the most commonly used. Abnormalities of the evoked potential indicate damage to the relevant pathway, either in the form of a conduction delay (increased latency) or reduced amplitude or both.
With the advent of magnetic resonance imaging (MRI). the use of evoked potentials is becoming much more restricted to specialized indications, such as providing a semi-objective measure of visual function.
Nerve conduction studies and electromyography
Using surface or needle electrodes, it is possible to record action potentials from nerves which lie close to the skin surface, as well as from muscles. If a nerve trunk is stimulated with a small electric potential, it is possible to record the resulting compound action potential (the sum of all the individual nerves’ action potentials) as it travels down the nerve. A normal compound action potential would have an amplitude of 5-30 microvolts, depending upon the nerve. If the recorded potential is smaller than expected, this provides evidence of a reduction in the overall number of functioning axons. Central conduction times can be measured using electromagnetic induction of action potential in the cortex or spinal cord by the local application of specialized coils.
Compound action potentials (CMAPs) can also be recorded over muscles in response to motor nerve stimulation. These are easier to record because the muscle amplifies the response, typical amplitudes being 1-20 millivolts. By measuring the response latency to stimulation of a nerve at two different points along its length, it is possible to calculate nerve conduction velocities (NCVs). This can be done for both sensory and motor nerves, and typical values are 50-60 mls. Slowing of conduction velocity is suggestive of peripheral nerve demyelination which may be either diffuse (as in a demyelinating peripheral neuropathy) or focal (as in pressure palsies or conduction block).
The principal use of nerve conduction studies is to identify damage to peripheral nerves and to determine whether the pathological process is focal or diffuse and whether the damage is principally axonal or demyelinating. It is also possible to obtain some information about nerve roots by more sophisticated analysis of responses to impulses initially conducted antidromically to the spinal cord, and then returning orthodromically to the stimulation point (F waves).
Fine concentric needle electrodes can be inserted into muscle bellies themselves and the potentials from individual motor units recorded. It is possible to record abnormal spontaneous activity arising from muscles at rest, such as fibrillations (a sign of denervation) or myotonic discharges. Abnormalities in the shape and size of muscle potentials can help in the differential diagnosis of denervation and structural muscle diseases. Myopathies caused by metabolic abnormalities (causing electromechanical dissociation rather than loss of fiber structure) show no changes on needle EMG.
Electromyography can also be used to investigate the neuromuscular junction. Repetitive stimulation of a nerve with trains of electrical impulses at 3-15/s does not normally result in a significant fall-off in the amplitude of the resulting muscle action potential. However, such a decrement is seen in myasthenia gravis and provides one of the key diagnostic features. Augmentation of the response to repetitive stimulation is seen in the Lambert-Eaton myasthenic syndrome, though usually at higher stimulation frequencies.
Imaging is crucial to the identification of lesions of the nervous system in disease. There are various techniques, based on the use of X-rays (plain radiographs, computed tomography (CT), myelography and angiography), magnetic resonance (MR imaging-MRI-or MR angiography), ultrasound (Doppler imaging of blood vessels), and radioisotopes (single photon emission computerized tomography (SPECT) and positron emission tomography (PET)). The application of various techniques depends upon the area of the neuraxis which is being investigated.
Head and orbit
The use of plain skull radiographs is largely restricted to the diagnosis of fractures and sinus disease, CT or MRI is needed to image pathology inside the skull. Which of these two scans is used depends on what information is being sought and. to some extent, how urgently it is required. as CT is often more easily available than MRI. CT will show bone and calcium well, and will easily image collections of blood. It will also detect abnormalities of the brain and ventricles, such as atrophy, tumors, cysts, abscesses, vascular lesions and hydrocephalus. Diagnostic yield is often improved by the use of intravenous contrast and spiral CT methods. It is, however, limited in its ability to image the posterior fossa (because of the surrounding bone density) and it is poor at detecting abnormalities of white matter and at allowing detailed analysis of grey matter.
MRI is much more useful in the investigation of posterior fossa disease as it is not affected by the surrounding bone. It is much more sensitive than CT to abnormalities of white and grey matter and is therefore useful in the investigation of inflammatory conditions such as multiple sclerosis and in investigating epilepsy. MRI can also provide additional information about structural brain lesions which may complement that available from CT. It is also useful in imaging the orbits, where special imaging sequences can be used to compensate for orbital fat and thereby allow clear views of extraocular muscles, optic nerve and other orbital structures.
Standard isotope brain scans are of little value in assessing structure if other imaging facilities are available. However, the blood flow and function of the cerebral hemispheres can be assessed by using either SPECT or PET.
Plain radiographs of the neck are useful in the investigation of structural damage to vertebrae, such as that resulting from trauma or inflammatory damage (e.g. rheumatoid arthritis). They can also provide implicit information about intervertebral disc disease, but not detailed information about the cervical cord or nerve roots, for which myelography or MRI is needed.
Myelography is invasive. Potential complications include headache, seizures and meningitis. With the advent of MRI its use is declining. Nevertheless, it is still of value if MRI is not available or the patient cannot tolerate lying within an MRI scanner. Radio-opaque contrast is injected into the lumbar theca and then moved up to the cervical region by tilting the patient. The contrast outlines the nerve roots and spinal cord, thereby providing information about abnormal structure.
Imaging of this region is similar to imaging the neck and plain radiographs are of limited use. Contrast can be injected into the lumbar thecal space and used to outline the lower nerve roots only (radiculography), or it can be run up to outline the conus and spinal cord (myelography). The information obtained may be enhanced by the additional use of CT following myelography. Non-contrast CT of the lumbar spine can be used to image the vertebrae and discs, but nervous tissue cannot be distinguished from CSF, although the addition of contrast by myelography allows individual nerve roots to be seen. As with the cervical spine, MRI provides a non-invasive way of obtaining high, resolution images of both the vertebral column and the relevant neural structures.
Various techniques are available to investigate extracranial and intracranial blood vessels. The least invasive is that of ultrasound (Doppler or duplex scanning), which is used to investigate the carotid and the vertebral arteries in the neck, usually as pan of the investigation of stroke. In skilled hands, reliable information can be provided about the degree of arterial stenosis and the technique often gives useful anatomical information, e.g. whether there is an ulcerating plaque. Information concerning the blood now in the intracerebral vessels is also becoming increasingly possible using transcranial Doppler. The anatomical resolution of Doppler imaging is limited and formal angiography may still be required. The latter is, however, invasive, and therefore carries a small but significant risk of stroke, or even death. Thus, the major role of Doppler imaging is as a screening test to determine whether invasive angiography is indicated.
Blood vessels can be outlined by the injection of radioopaque contrast. The X-ray images obtained can be enhanced by the use of computer-assisted digital subtraction or by the use of spiral CT. Contrast may be injected intravenously or intra-arterially. The former requires a much higher total dose of contrast and the images obtained are not as good, but the latter involves feeding catheters up through the arterial tree and is thus associated with a higher complication rate. Formal intra-arterial angiography is required to delineate lesions of the extracranial carotid artery prior to endarterectomy, and is also used to investigate abnormalities of intracerebral vessels such as arterial (berry) aneurysms or arteriovenous malformations or to delineate the blood supply of tumors prior to surgery.
Flowing blood can be detected by specialized MR sequences in MR angiography. The anatomical resolution is still not comparable to that of intra-arterial angiography, but the investigation is non-invasive.
Many systemic conditions affect the nervous system and these can often be diagnosed with the help of blood tests: for example, confusion due to hypothyroidism, a stroke due to systemic lupus erythematosus, ataxia due to vitamin B12 deficiency, or myelopathy due to syphilis. The blood tests relating to general medical conditions which affect the nervous system are dealt with in the sections dealing with the conditions themselves.
There are, however, a number of blood tests which are used in investigating specific neurological diseases. These include hematological tests (e.g. looking for acanthocytes to diagnose neurocanthocytosis), biochemical tests (e.g. creatine kinase in muscle diseases, copper studies to diagnose Wilson’s disease) or tests to help diagnose innumerable infections of the nervous system. In addition, there are a number of specific antibodies which are useful diagnostically. These include antibodies to acetylcholine receptors and skeletal muscle, seen in myasthenia gravis and to voltage-gated calcium channels in Lambert-Eaton myasthenic syndrome. Antibodies to different types of ganglioside (glycoproteins expressed on nerve membranes) can be seen in various types of neuropathy including multi focal motor neuronopathy, and the Guillain-Barre syndrome (particularly the Miller Fisher variant). Also, antineuronal antibodies provide markers of paraneoplastic cerebellar or neuropathic syndromes.
An increasing number of inherited neurological conditions can now be diagnosed by DNA analysis. These include diseases caused by increased numbers of trinucleotide repeats, such as Huntington’s disease, myotonic dystrophy and some types of spinocerebellar ataxia. Also, defects of mitochondrial DNA can be detected in many conditions including Leber’s hereditary optic neuropathy and some syndromes causing epilepsy or stroke-like syndromes.
This involves the insertion of a needle between lumbar spinous processes, through the dura and into the cerebrospinal fluid (CSF) under local anaesthetic. The intracranial pressure can be measured and CSF removed for analysis. CSF is normally clear and colorless. Tests usually performed on CSF include centrifuging to determine the color of the supernatant (which is yellow or xanthochromic some hours after subarachnoid hemorrhage), biochemistry (glucose, total protein, and protein electrophoresis to detect oligoclonal bands), microbiology, immunology (e.g. Venereal disease reference laboratory (VDRL), paraneoplastic antibodies) and cytology (to detect malignant cells).
Lumbar puncture is indicated in the investigation of infections (e.g. meningitis or encephalitis), subarachnoid hemorrhage, inflammatory conditions (e.g. multiple sclerosis, sarcoidosis and cerebral lupus) and some neurological malignancies (e.g. carcinomatous meningitis, lymphoma and leukemia), and to measure CSF pressure (e.g. in idiopathic intracranial hypertension). It is, of course, part of the procedure of myelography, and can be part of therapeutic procedures, either to lower CSF pressure, or to administer drugs.
If there is a space-occupying lesion in the head, lumbar puncture can result in a shift of intracerebral contents downwards, towards and into the spinal canal. This process is known as coning, and is potentially fatal. Consequently, lumbar puncture is contraindicated if there is any suggestion of raised intracranial pressure (e.g. papilloedema), depressed level of consciousness, or focal neurological signs suggesting a cerebral lesion, until imaging of the head (by CT or MRI) has excluded a space occupying lesion or hydrocephalus.
About 30% of lumbar punctures are followed by low pressure headache, which can be severe. Other minor complications involve transient radicular pain during the procedure, and pain over the lumbar region. Provided the test is performed under sterile conditions, infections such as meningitis are extremely rare. Lumbar puncture is contraindicated in anticoagulated patients but not in those on aspirin.
Nerve and muscle are occasionally biopsied to assist in the diagnosis and management of a number of neurological conditions. Likewise, it is occasionally necessary to biopsy brain or meninges.
Nerve is biopsied as part of the investigation of peripheral neuropathies. Usually, the sural nerve is sampled at the ankle or the radial nerve at the wrist. Histology is often able to help identify underlying causes in demyelinating neuropathies (e.g. vasculitic) or, occasionally, infiltration with abnormal substances such as amyloid. However, nerve biopsy is not performed unless it is reasonably likely to diagnose a potentially treatable condition such as an inflammatory neuropathy, since there is an appreciable morbidity risk.
Skeletal muscle biopsy is performed more frequently. The quadriceps muscle is often sampled, though this depends somewhat on which muscles are affected. Indications include the investigation of primary muscle disease, as muscle histology can be used to distinguish neurogenic wasting, myositis and myopathy, which may be difficult to distinguish clinically. Histology and enzyme histochemistry can also be helpful in the diagnosis of more widespread metabolic disorders, such as mitochondrial and some storage diseases. Though pain and infection can follow the procedure, these are much less of a problem than following nerve biopsy.
The nature of lesions demonstrated by brain imaging can often be inferred from the appearances as well as the history, examination and other, less invasive investigations. However, there are situations in which the nature of lesions is not clear and it is important to obtain tissue for histological examination. Likewise, it is sometimes necessary to biopsy the brain parenchyma itself in unexplained degenerative diseases (e.g. unusual dementias) so as not to miss potentially treatable disease.
Brain biopsy used to require full craniotomy. However, owing to the increased availability and sophistication of cerebral imaging, it is now possible to biopsy most lesions stereotactically through a burr hole in the skull. The complication rate of such stereotactic biopsies is much lower than that of open craniotomy, but hemorrhage, infection and death still occur. Hence, brain biopsy is only considered if diagnosis cannot be reached in any other way.