APPROACH: WHERE IS THE LESION, WHAT IS THE LESION?
Exam - localising with clinical signs
There are so many examination manoeuvres that there is no such thing as a ‘complete neurological examination’ - it is tailored to the problem being assessed.
I can’t list every situation or manoeuvre and it is not my aim - there are entire books on this. The cases will illustrate its utility in given situations. This section is an overview of its purpose.
The exam also serves several purposes, and is not only about localisation but also severity grading. It is also useful to look for clues to an underlying aetiology, not only at the lesion itself. Finally it is an important means to build rapport and confidence. Again, my aim here is to focus on localisation.
When we examine we already should have one or several hypotheses in mind, and we look for evidence to confirm or refute these - ideally, ‘narrowing it down’ to one or a few likely explanations we may then use investigations to confirm/refute. I cannot stress enough the point that bypassing this clinically process and simply doing tests is not only inefficient but also causes all sorts of issues - summarised elsewhere .
For example - if we are seeing a patient with bilateral leg weakness sparing the arms and cranial segment, the differential includes the following:
The most useful way to split these is reflexes, as if they are brisk it would suggest a UMN lesion, and their absence would point to an LMN issue (bar in the hyperacute phase of LMN lesions). They may be normal however. From there we can then look for clues that distinguish the other lesion sites listed above, for example sphincter disturbance (cauda equina), co-existing sensory loss in a length-dependent manner (neuropathy), fatiguability (NMJ) or proximal predominance with no sensory features (myopathy).
However, sometimes we may not know exactly what the deficit is based on the history, if multiple options exist, so the exam can help this. For example, a patient with difficulty using a hand may have weakness, numbness, ataxia, apraxia or something else - the exam can then identify the true cause.
It is sensible to have a standard structure for most situations and to be able to customise to the problem. A routine examination should include cranial nerves, limbs (tone, power, reflexes, coordination and basic touch sensation), gait and stance. A general sense of cognition should be apparent through the history and the patient’s ability to follow instructions.
Depending on the situation, we may include other elements which are not ‘routine’. Examples include:
Some examination sequences are standardised and are used to grade severity with an overall score, which influences treatment. One of the most commonly used is the National Institute for Stroke Score (NIHSS), which allows rapid screening for major deficits due to stroke to enable decision-making around reperfusion therapies. It includes tests for major weakness in the proximal muscles in each limb affecting anti-gravity strength, rather than looking at individual muscles in depth, and it ignores things which have no relevance in hyperacute stroke (such as reflexes).
In contrast, examination of a suspected peripheral problem can require a detailed examination of muscles to map the lesion precisely. If doing this for an upper limb issue such as difficulty with fine movements in one hand, the examiner would carefully examine the affected region to help find the cause - for example ulnar neuropathy, T1 radiculopathy or even a motor homunculus lesion. Tests and levels of detail chosen are tailored to the given problem.
In addition to localising the lesion we may also use the exam to look for causes for a suspected disorder, for example hypertension or cardiac abnormalities in a suspected stroke. We may also look for additional clues that co-occur in a condition, including non-neurological ones such as eye or skin abnormalities. There is lots to say on this, but it's less about 'Where?' and more 'What?'.
We will look at some specific situations in more detail now.
Sensory lossSensory complaints are very common and many reflect benign causes, particularly those with positive symptoms alone (tingling), which migrate, and have no sensory loss associated. However, serious disorders can produce persistent sensory disturbance - including any of the lesion types in this resource. Sensory problems can be challenging to assess.
The most important task is mapping the affected region and measuring involved modalities. From there we can consider possible anatomical explanations. These follow rules outlined in the Where? section.
Brain lesions often affect the contralateral hemi-head and body, although can be more focal if in the sensory cortex (unless large). Cortical lesions may also produce higher-order sensory abnormalities such as agraphaesthesia and astereognosis even if primary sensation is spared.
Thalamic lesions can produce a variety of patterns, including complete hemihead/-body ‘splitting’ but also more unusual distributions such as cheiro-oral syndrome
Brainstem lesions tend to cause crossed signs, with ipsilateral cranial region sensory loss, and contralateral body sensory loss. The modalities involved depend on the lesion; for example lateral medullary lesions affect the spinal trigeminal and spinothalamic tracts, so alter pain and temperature sensation but spare other modalities.
Cord lesions vary with the lesion level and also side, site and size in the cord, hence the modalities involved depend on this, and patterns vary, as was discussed in Where?. Finding the sensory level is important, and the lesion is typically at or slightly higher than this. If the arms are involved the lesion cannot be lower than the high thoracic cord, although some cervical lesions do spare the arms - so can be missed if we omit this region on imaging in someone who only has lower limb (and perhaps torso) symptoms.
Lesions in the periphery vary.
Polyneuropathy features glove-and-strocking sensory loss, so if the sensory loss goes to the thighs and spares the hands this suggests a more central cause.
Root, plexus and mononeuropathy lesions will affect a specific distribution in a region of skin, so we need to map this out. These regions may overlap (for example L5 and common peroneal nerve), and it can be difficult to be sure of the lesion site, so other features such as weakness (if present) help this.
The most difficult lesions to site involve smaller areas. It is generally easier the more extensive the sensory loss, e.g. all 4 limbs or entire hemibody. One presentation which is difficult is sensory loss to the mid-shins, which can be polyneuropathy or lesion in the cauda equina or cord - and there may not be adequate additional evidence to know which the problem is on clinical grounds alone.
WeaknessThe approach taken to localising weakness is similar to sensation, and the key is first mapping affected regions - a lesion may be anywhere in the motor tract between its cortical origins and target muscle groups. The same patterns apply as in the sensory section, for example partial or total head and body weakness on the opposite side with brain lesions - with the exception that some areas have bilateral UMN supply so are not weakened by a unilateral lesion (movements in the forehead and jaw). Otherwise, the same is true with the brainstem (crossed signs), spine and other regions.
Other features that help are tone and reflexes, in terms of separating UMN versus LMN, though these can take some time to develop. They are helpful if seeing someone several days or weeks into a problem, but not necessarily discriminatory in the first days. An anterior spinal artery infarction produces flaccid areflexic paraplegia at onset - spasticity comes later.
Sensory features, if also present, can be helpful clues. In the above example of a spinal infarct there is generally loss of spinothalamic modalities bilaterally but sparing of the dorsal column functions, and there should be a sensory level in the trunk - this helps differentiate from other causes of flaccid paraplegia such as cauda equina syndrome.
Other clues from distribution were explored in Where? - such as prominent ocular, facial and bulbar features in NMJ disorders, proximal weakness in myopathy, and distal predominance in neuropathy.
Visual problemsThe visual pathways start in the eye and work back through the optic nerves, chiasm, optic tracts, thalamic synapse, then the two radiations until terminating in the occipital cortex. What starts as a single eye having representation of bilateral fields including upper and lower halves is then divided, with one half of the visual space from both eyes ending up in the opposite occipital lobe, and the upper and lower quarters of vision represented either side of the calcarine sulcus.
This is primary vision. Lesions anywhere in this pathway lead to visual loss, and the amount lost and from what area of vision depends on where the lesion is.
There is then a complicated system of higher-order visual processing distributed between the occipital, parietal and temporal lobes. Lesions here cause more complicated effects rather than primary visual loss. Examples include problems perceiving movement (akinetopsia), or recognising faces (prosopagnosia) or reading words (alexia) despite intact vision. Small, focal lesions can produce isolated deficits such as this.
When we examine our task is to find out what the problem is and where the pathology may lie. Our assessment is less detailed or instrument-guided than what ophthalmologists do, but with very little equipment we can test acuity, fields, examine the optic disc, as well as the pupillary light and accommodation reflexes and look for colour vision. Examining the blind spot is also helpful. Note that problems such as nystagmus may also produce ‘blurred’ vision due to constantly disrupted foveal fixation, and diplopia (e.g. due to a VI palsy) may be reported as ‘blurred’ vision too.
All of this requires good technique to minimise risks of artefacts. The patient must be clear what they are being asked to do, and we need to clarify what they are experiencing if an abnormality is noted, too.
If the patient is confused, dysphasic, agitated, drowsy or otherwise has altered ability to interact, it becomes difficult to meaningfully assess vision and we resort to basic measures such as testing each hemifield with a threat stimulus (a hand quickly brought towards their face) to see whether they blink - a common test in hyperacute stroke patients.
VertigoVertigo is common. Localisation is broadly split into the vestibular system (peripheral) and its components within the brain and cerebellum (central).
If someone has persistent vertigo, if there are obvious features of a central lesion such as hemi-ataxia or sensory loss then localisation to a central cause is straightforward, and we then localise it further using the principles described earlier, e.g. the site in the brainstem.
If such features are not obvious then we have to look in more detail. Eye signs are the most informative, and if there is nystagmus, its characteristics can discriminate. Peripheral nystagmus is horizontal and may have a torsional component, and the direction of the horizontal fast phase doesn’t change. Anything else is central. We can also test the vestibulo-ocular reflex (VOR) by head impulse testing to see if the eyes can fix a target despite rapid head movements. If altered the problem is peripheral - which is reassuring. People with peripheral vertigo can also usually walk unaided, even if it feels unpleasant - if they can’t the cause is usually central.
In episodic vertigo the patient may be asymptomatic during the assessment, so instead we use provocative manoeuvres to see if we can trigger an attack of BPPV. These test the individual semicircular canals and allow localisation of the cause, and we can follow up with a corrective manouevre.
Cognitive problemsAs written above, cognitive assessment instruments test the different faculties and this is part of localisation. We can use these in our examination in addition to other tests and observations.
We can formally test cognitive functions using psychological tests, and various instruments combine these to give quantitative scores. The different cognitive domains localise to areas of the brain. However, even without these we can assess cognition and localise the brain regions involved during the examination. Even before this, cognitive performance is evident during the history, and we may spot lots of issues with language, memory, attention and organisation of thought processes.
Frontal features include problems with social cognition, and we may note things such as an air of lack of interest in the consultation, an inappropriately jolly or joking manner, and a tendency to interrupt and make insensitive remarks (often this is evident watching the reactions of the accompanying person present). People often struggle to keep up with the flow of the assessment and may appear to get stuck on a subject or task that we have moved on from (perseveration).
In addition, utilisation behaviour may be evident - the patient picks up items around the room and starts using them without any instructions to do so, for example a pen and paper on the desk.
Frontal release signs are primitive features that emerge with frontal lobe damage. They include the pout, root, snout, grasp and palmomental reflexes.
Patients with frontal lobe dysfunction struggle with abstraction, such as explaining the meaning of proverbs or similarities between discrete objects (e.g. knowing that a banana and orange are both fruit). They also struggle with timed word fluency, despite having fluent spontaneous speech and no dysphasia - because this demands the ability to concentrate and think under pressure.
Posterior brain cognitive functions include numeracy and visuospatial functions. We can ask people to do simple arithmetic, or to draw objects such as a flower or clock. There may be problems with praxis, for example copying gestures or miming an action most people would know (turning a key in a lock, brushing teeth, putting on a hat). Visual or sensory inattention may also be evident in parietal disease so is worth testing, likewise checking for a cortical sensory deficit.
Memory is important to test. We can ask about recent events in the news, or the patient’s overall awareness of their illness. We can also give information to memorise such as 3 items (lemon - key - ball) or an address. We test immediate recall (‘repeat after me…’) and delayed recall (after an interval).
A problem with immediate recall does not necessarily imply a memory issue - the patient may have problems with attention. However, intact immediate recall and poor delayed recall implies a problem with processing and storage of new information, a memory issue. This localises to the hippocampus and its outputs via the fornix, mammillary bodies, thalamus and cingulate (the Papez circuit). It is very obvious when assessing someone mid-attack of transient global amnesia, when several minutes later they not only can’t remember the items listed but that there was even such a memory task given. This is a benign but alarming disorder - the problem resolves in under 24 hours and is usually not recurrent.
Given many neurological conditions can involve cognition as part of a wider set of problems, it is sensible to examine reasonably broadly in cognitive presentations, for example checking for parkinsonian features, cerebellar signs and features such as a startle response. This is less about looking for a specific localisation of cognitive functions, and more for a clue to a particular neurological disorder that underlies the cognitive presentation.
Putting it all together...Having taken a careful history, thought about the symptoms and what site and process they may map to, then examined to support our suspicions - we can now attempt a clinical formulation of the patient's problem.
Clinical Formulation