This chapter is relevant to the aims of Section K1(v) from the 2017 CICM Primary Syllabus, which asks the exam candidate to "describe the major sensory and motor pathways (including anatomy)" up to an "L2" (vague awareness) level of understanding.
What does that look like when you're answering questions about it? Spinal cord anatomy. Specifically, the effects of spinal cord transection. Only two questions from (very old) primary exam papers had asked about the effects of spinal cord injury at varying levels. But then spinal injury has been a common topic in the Second Part exam, where it has appeared numerous times, in countless permutations. Behold, the puzzling distribution of spinal questions:
|First Part Exam||Second Part Exam|
As such, it is unclear how the examiners have blueprinted this assessment process to the overall pattern of trainee development. Where during their larval stage do they think it belongs? If assessment drives learning, then this is a clear signal that they want you to leave it until your fellowship exam, when you are close to the end of your training. Unfortunately, by that stage you will no longer be interested in the finer details of white matter tract anatomy, and will just want to cynically dissect a bunch of negative trials. So, striking while the iron is still full of curiosity and enthusiasm, this chapter is something of a deep dive into spinal neurology, flavoured in places with pointless trivia.
But, in case there is no time for that:
- Major sensory and motor pathways:
- White matter tracts, long bundles of axons, whereas the cell bodies reside in the grey matter.
- Many decussate, i.e cross midline from their origins to their destination
- Many are made up of three or more neurons
- Motor neuron pathways:
- First order neurons: motor cortex, "upper motor neurons"
- Second order neurons: grey matter of the spine, "internuncial" neurons
- Third order neurons: grey matter of the spine, "lower motor neurons"
- Sensory neuron pathways:
- First order neuron cell bodies are in the dorsal root ganglia
- Second order neurons are in the dorsal horn of the spinal cord
- Third order neurons are in the destination organ, eg. thalamus
- Main motor tracts of the spinal cord and their function:
- Septomarginal fasciculus and interfascicular fasciculus: internal spinal reflex arcs
- Lateral corticospinal tract: fine quick voluntary movement
- Lateral reticulospinal tract: posture, flexor movements
- Rubrospinal tract: posture, flexor movements
- Anterior corticospinal tract: coarse voluntary movement
- Anterior reticulospinal tract: posture, extensor movements
- Vestibulospinal tract: posture, extensor movements
- Tectospinal tract: reflex postural movements (visual stimuli)
- Main sensory tracts of the spinal cord and their function
- Dorsal column tracts: propriception, vibration, light tough
- Lateral spinothalamic tract: pain and temperature
- Posterior spinocerebellar tract: tendon and joint position
- Anterior spinocerebellar tract: tendon and joint position
- Spinoolivary tract: cutaneous and proprioceptive information
- Spinotectal tract: afferent information necessary for the movement of the head in response to painful stimul
- Anterior spinothalamic tract: coarse touch and pressure
Reading through the mess below, one may be overwhelmed by the irresistible urge to study from something more professional. Unfortunately, though there are multiple textbooks that claim to present this information "Made Easy" or "Ridiculously Simple", there are few peer-reviewed journal articles on the topic, and basically none of them are available for free. If you're made of money and insist on buying a neuroanatomy textbook at some stage, for whatever reason, make it Snell's Clinical Neuroanatomy by Splittgerber (mine is 2018). Unless stated otherwise, the rest of this chapter is a summary of Snell's. And if the pragmatic reader is really only trying to answer past paper questions (which mainly asked about cord section or hemisection), Diaz and Morales (2018) have everything required.
When the college ask you to describe "the major sensory and motor pathways (including anatomy)", they are referring to the white matter tracts of the central nervous system. A "tract" is just a word, perhaps misapplied (in the sense that it is a stretch of white matter, in the same sense as a tract of woodland or a length of text), but"tract" and "pathway" are probably the better terms we've got for these structures. Other, weirder nomenclature had been popular historicall, which occasionally persists in neuroanatomy textbooks because all anatomical sciences suffer from a severe infestation of Latin words and eponyms. For example, the venerable Core Text of Neuroanatomy by Carpenter (1978) refers to fasciculi and funiculi, a funiculus being a bundle of fasciculi.
Anyway. Unlike the design of a circuit board, the normal logic which guides the layout of connecting tracks is lost in the brain, because it develops from a pile of goo. If you can think of a sensible way to run a circuit, you can guarantee that you won't find that in the CNS. For example, decussation (the crossing over of white matter fibres from left to right) is almost totally unique to vertebrates, and nobody can explain why it is like that, what purpose (if any) it serves, and why specific fibres cross at specific points whereas others do not. The prevailing view right now (the "somatic twist hypothesis") patiently explains that you are basically just an upside-down crab, rotated 180° somewhere around the oropharynx.
But why? Can the ipsilateral pain sensors not just inform the ipsilateral hemisphere? Would that not be simpler? One reasoned explanation for why decussation might have a functional role (rather than being some freakish accident of crab-twisting evolution) has been offered by Ramón y Cajal, whose 19th-century work was beautifully reviewed in Mora et al (2019). In short, if there was no decussation, the cortical representation of bilateral stimuli would be incomplete. Each hemisphere would receive its half of the representation, and it would actually be more computationally difficult to integrate the inputs into a working whole. Cajal explained himself by using the optic system, but his theory seemed to work for tactile and auditory stimuli as well. The original diagrams from his 1898 papers are sufficiently awesome to be reproduced here without any permission whatsoever:
However, this is not a widely accepted explanation, and as an explanation it has many holes. Behold, the aforementioned crab. It has a set of eyes that do not just face forward and observe the same scene- in fact they are on stalks, and could point in any damn direction they please. Surely the cortical mapping of such complex visual data would require decussating fibres? But in fact invertebrates do not have decussation as the general rule, i.e. its obviously not a necessary part of having a nervous system that accurately maps your geographic surrounding. In fact you can have a very complex nervous system and perfect spatial comprehension with a unilateral chiasm (Braitenberg, 1965 points out that cephalopods don't have decussating optic fibres). In short, we have no idea. And don't even start on the topic of hemispheric dominance.
The ICU trainee will (hopefully) never look directly upon brain tissue, and so for them the most relevant representation of neuroanatomy will probably be a sectional radiological approach. For this, the best sort of reference would probably be something like Wucoco et al (2013), "White matter anatomy: what the radiologist needs to know", assuming that's also representative of what the intensivist needs to know. Again, "needs" is a strong, strong statement when it comes to the neuroanatomy of white matter tracts. It would be unexpected for a detailed question on this to ever show up in any of the CICM exams. As a compromise between not saying anything at all and saying too much, this table from Wucoco is offered here to stimulate interest.
|Cingulum||Cingulate gyrus to the entorhinal cortex||Affect, visceromotor control;
response selection in skeletomotor
control; visuospatial processing
and memory access
|Fornix||Hippocampus and the septal area to hypothalamus||Part of the Papez circuit; critical in
formation of memory; damage or
disease resulting in anterograde
|Superior longitudinal fasciculus||Frontotemporal and frontoparietal
|Integration of auditory and speech
|Ipsilateral temporal and occipital
|Visual emotion and visual memory|
|Frontal lobe to ipsilateral parietal
lobe—name being a misnomer
|Spatial awareness, symmetric processing|
|Ipsilateral frontal and occipital,
posterior parietal and temporal
|Integration of auditory and visual association cortices with prefrontal cortex|
|Uncinate fasciculus||Frontal and temporal lobes||Auditory verbal and declarative memory|
|Thalamic radiations||Lateral thalamic nuclei to cerebral
cortex through internal capsule
|Relay sensory and motor data to precentral and postcentral cortex|
|Motor cortex and cerebral peduncle
through internal capsule
|Descending motor fibers from primary motor cortex, ventral and dorsal premotor areas, and supplementary motor areas|
|Corpus callosum||Corresponding cortical areas of both
|Interhemispheric sensorimotor and auditory connectivity|
|Anterior commissure||Olfactory bulbs and nuclei and
|Integral part of the neospinothalamic tract for nociception and pain sensation|
Any discussion of long white matter tracts always needs a crossectional diagram of the spinal cord. These come in a variety of shapes and sizes, and the best ones tend to have coloured coding for motor and sensory pathways, as well as labels that relate the structure to its function. Disappointment with official diagrams has led to the creation of the one below.
These tracts can also be anatomically organised into funiculi, i.e. there is a dorsal, ventral and lateral funiculus, but this system of organisation does not have a lot of functional merit. For example, the lateral funiculus contains the sensory spinocerebellar spinothalamic spinoolivary and spinotectal tracts, as well as some important motor tracts (lateral corticospinal, lateral reticulospinal and the rubrospinal tracts)- none of which are especially related to one another.
This crossectional diagram fails somewhat, insofar as it is unable to describe the course of the aforementioned tracts. Fortunately, in a refreshing twist uncharacteristic of neuroanatomy (and of anatomy in general), the tracts are named sensibly, i.e their name relates to the things they connect. The spinothalamic tract travels to the spine, from the thalamus. The lateral corticospinal tract originates from the cortex, and is positioned laterally. You really can't ask anything more from a nomenclature.
Trying to organise these into funiculi has basically failed in the process of writing about them, and so the list of tracts here is really just thrown together in whatever way seemed to flow the best, as far as explaining their function is concerned. Which means that this list is not ordered according to importance or exam relevance. Sometimes, a list is just a list.
Now, α motor neurons are listed as the lower motor neurons for most of the spine section images below, but really they are not the only neurons down there. The α motor neurons
There are also γ motor neurons
All of these pathways have common stereotypical pattern, though there are exceptions.
The pathways might all be totally different, but there is only one set of muscles for them to control, and so it would make absolutely no sense for each individual tract to send their fibres directly to the muscle cells involved. Instead, all first and second order neurons from all the different tracts synapse with the same lower motor neuron, which then acts as the final common pathway for the signal.
Lateral corticospinal tract connects the upper motor neurons in the pre-central gyrus with muscles. The descending motor fibres decussate at the level of the medulla and continue along the contralateral corticospinal tract. At the level of their nerve root, these fibres synapse with some "internuncial" neurons, or just directly with the α motor neurons of the spinal grey matter. These cholinergic lower motor neurons then innervate muscle.
Specifically, the corticospinal tracts are involved in voluntary movements which require rapid and precise control of muscles. This ranged from watchmaking to catching a football. The axons are thick, well myelinated, and have excellent conduction velocity. Ingram et al (1985) measured something like 67 m/s, and numbers up to 100 m/s have been reported.
Anterior corticospinal tract is basically the same upper motor neurons sending projections to a slightly different group of muscles. The fibres do not decussate with the rest of the big motor bundles in the pyramids -they cross at the level of the spinal nerve. The lower motor neurons from here seem to innervate mainly muscles of the torso, whereas the corticospinal tract is more interested in the limbs.
The upshot of this separation of fibres is not entire clear, from a clinical perspective. Surely this should mean that some lesions take out the limbs but spare the trunk? This is a thin and unimportant white matter structure, making it hard to study (and isolated lesions are relatively rare). However, it does have importance to motor control. For example, contralateral stroke in the brainstem can destroy the descending motor fibres after they have decussated, but the anterior corticospinal fibres would remain ipsilateral and would therefore be spared, leaving the patient with some ability to coordinate their trunk and proximal muscles. Jang & Kwon (2013) report on just such a case, remarking that the hemiparesis got worse after the anterior corticospinal tract fibres were also affected by stroke.
Anterior reticulospinal tract, otherwise known as the pontine reticulospinal tract, contains fibres which originate from the nuclei of the pontine reticular formation. These fibres innervate the skeletal muscles of the trunk and the extensor muscles of the upper limbs. Interestingly the upper motor neurons of the cortex exert only the most minimal influence on the activities of this tract. As far as one is able to establish from pixellated textbook scans and random blogs, most of these fibres do not appear to cross the midline.
The lateral reticulospinal tract, otherwise known as the medullary reticulospinal tract, contains fibres which originate from the medulla. They have what is described by Netter as a "flexor bias". Unlike the anterior (pontine) tract, the medullary upper motor neurons are heavily influenced by cortical input. Also they decussate, whereas the pontine reticulospinal fibres stay ipsilateral.
Together, the reticulospinal tracts serve as the levers of motor control for coarse involuntary movement of large limb muscle groups, and the overall tone of these muscles. The underlying reason for their activity is the premise that in order for the movement of a joint to occur, when one group of muscles voluntarily contracts another group must relax. Additionally, certain continuous or repetitive movements (eg. the constant minor corrections of tone required to maintain upright posture) probably do not need constant cortical control and can be outsourced to simpler circuits.
Thus, the reticulospinal tracts run an automated background cron job which keeps the muscle tone normal. Both tracts tend to balance each other, with the result that tonic flexion and tonic extension are basically equal. Damage to the cortex, and loss of input into the medullary nuclei which control the lateral reticulospinal tract, results in a loss of the medullary "flexor bias", and the pontine nuclei start to dominate the flexor-extensor balance. This contributes to the neurological "posturing" findings of the brain-injured patient whose cortex is for whatever reason disabled.
Rubrospinal tract is the next logical tract to discuss, now that we're on to the topic of brain-injured posturing. This tract originates from the red nucleus, a little structure in the medial rostral tegmentum of the midbrain. Fibres from here decussate at the level of the brainstem and then travel down the spine to innervate mainly flexor muscles, like the lateral (medullary) corticospinal tract.
The neurons of the red nucleus receive fibres from the spinothalamic tract before it ascends to the thalamus. The result is a shortcut for pain stimuli to affect the motor output of the red nucleus. It is theorised this flexor reflex produces a rapid and involuntary withdrawal from a painful stimulus which is present mainly in the upper limbs. The descending inhibition of this reflex normally keeps it in check, but with damage to the cerebral hemispheres the red nucleus is left to do as it please. The result is a stereotypical flexor posture, described as an M3 score by the Glasgow Coma Scale (Riddle et al, 2009).
Injuries below the red nucleus (i.e. anything involving the midbrain) produce a situation where this flexor input is totally lost. At the same time, cortical input into the medullary "flexor-biased" tone controllers is lost. With only pontine reticulospinal and vestibulospinal nuclei remaining, motor responses begin to heavily favour the extensor muscle groups, with the result that painful stimuli produce an extension of the elbows, internal rotation of the shoulders, extension of the knees and plantarflexion of the ankles. This stereotypical posture is scored an M2 score on the GCS, and is associated with poorer outcomes, as it typically means that brainstem structures are injured (which is somehow even more horrible than just straightforward cortical brain damage).
Vestibulospinal tract is the other white matter strand that contributes to the extensor posturing of severe brain injury. It stretches down from the vestibular nucleus of the medulla. The normal function of this tract involves the control of extensor muscles, which contributes to the maintenance of posture.
Tectospinal tract stretches out of the superior colliculus of the midbrain. These fibres decussate early, high up in the brainstem, and carry on through the anterior white matter of the spinal cord. They really only seem to be present in the upper spinal cord and their function is summarised by Snell's as "reflex postural movements in response to visual stimuli".
Septomarginal fasciculus and interfascicular fasciculus are little dorsal motor tracts which are barely mentioned in the anatomical literature, and often omitted by medical textbooks. When one is able to find any reference of them, the function attributed to them is the carriage of fibres specifically concerned with spinal reflex arcs. This means these fibres should not actually ascend all the way into the brain - they are for local spine-to-spine communication only. However, no good descriptions of their anatomical position or neural connections are available, and so no diagram is available to describe them.
The ascending pathways have a predictable organisation:
Or, in the form of art:
This is just a representative example: specific sensory pathways may not do exactly this (there may be more neurons involved, and they may directly connect to motor neurons, as in the case of reflex arcs). However, one cannot help but notice that this diagram above is also seen in a million textbooks, and is therefore sufficiently likely to catch the eye of a tired examiner on the night before their viva script is due. There is a non-zero chance that the trainee may at some stage have to reproduce this neuron diagram.
In contrast, the chance that the sensory tracts will need to be discussed independently or in any great detail is much closer to zero. However, for completeness:
Dorsal column tract: fasciculus gracilis and fasciculus cuneatus are the two main columns, and they convey information about proprioception light touch and vibration. The gracilis carries fibres with information from the lower limbs, and you can remember this easily because the gracilis muscle is in the leg.
The first order neurons for these tracts don't synapse in the dorsal horn as would be expected. Instead, they ascend along the dorsal white matter columns until they synapse with second order neurons in medullary dorsal column nuclei (also called gracilis and cuneatus). Only after this do the fibres decussate as the "internal arcuate fibres", to synapse with the third-order neurons in the thalamus.
Lateral spinothalamic tract carries pain and temperature sensation. Sensory axons from first-order neurons travel another level up the spinal cord via the tract of Lissauer before synapsing with their second order neurons in the dorsal horn (apparently using Substance P as their neurotransmitter). From there, the destination is third order neurons in the thalamus (specifically the ventral posterolateral nucleus), and then the sensory cortex.
In the tract itself, the ascending fibres are layered, so that the innermost medial fibres are from the highest part of the cord. The outermost fibres are therefore the lowest, i.e. from the sacrum and the legs. This weird factoid has relevance: when the central canal is swelled by a syrinx, the medial pain and temperature fibres from the upper limb take the first hit, and sacral and lower limb sensation may be spared, giving rise to a "cape-like distribution" of sensory loss and paraesthesia.
Anterior spinothalamic tract is organised in much the same way (i.e. the most lateral parts are from the lowest spinal levels). It transmits coarse touch and pressure information, also to the thalamus.
The spinothalamic tracts fibres decussate across the midline at the level of the spinal cord, instead of waiting to get up to the brainstem like the rest of the spinal long white matter structures. Why this happens is unclear. The decussation was originally discovered by Charles-Édouard Brown-Séquard in 1846, and since then nobody has really been able to offer a satisfactory explanation, not that any answer to that question would ever be anything other then pure speculation.
Posterior and anterior spinocerebellar tracts are even weirder. The first order neurons synapse with their targets at the base of the dorsal horn, but then the tracts split into two. The joint position sense and muscle tension/tendon stretch information from the lower limbs and trunk ascends in the ipsilateral posterior spinocerebellar tract and up into the ipsilateral cerebellar hemisphere. Some of the same information, as well as some additional tidbits from forgotten musculoskeletal elements like fascia and skin, instead crosses the midline at the same spinal level and ascends as the anterior spinocerebellar tract. Then, once they ascend into the cerebellum through the superior peduncle, some of these fibres cross the midline again, back to the ipsilateral cerebellar hemisphere. Most readers will sure agree that there can't possibly be any point to this.
Spinoolivary tract carries information which Snell's describes as coming from "cutaneous and proprioceptive organs". It does not seem like a very important tract, in the sense that it is very small and unlikely to be damaged on its own. Again, for some reason there are two decussations here (across midline and then back again), except this time the second crossing is performed by their order neurons arising from the medullary olivary nuclei.
Spinotectal tract runs close to the spinoolivary tract, but unlike the latter it does not cross back to the ipsilateral brain. Instead, the second order neurons synapse with the superior colliculus of the contralateral midbrain. This tract carries afferent information necessary for the spinovisual or "tectospinal" reflexes which coordinate the movement of the head in response to painful stimuli. Some of the fibres also synapse in the periaqueductal grey matter, where the tract also plays some sort of role in pain gating. Textbooks and online resources mention some kind of inhibitory influence on pain transmission.
As promised, here are some links to the spinal chapters from the Fellowship Exam preparation resources:
Annoyingly, the college has asked about spinal cord injury syndromes in both the First Part exam and in the Second Part Exam. To simplify revision and annoy the Google algorithm, the exact same information will be reproduced here:
Anterior cord injury
Posterior cord injury
Central cord syndrome
Conus medullaris syndrome