What is Dr. Zubair Khalid's research focus?

Dr. Zubair Khalid specializes in molecular virology, mRNA vaccine development, and computational biology, with a focus on avian pathogens like IBDV and Avian Reovirus.

Where is Dr. Zubair Khalid currently working?

Dr. Zubair Khalid is a Postdoctoral Research Associate at the University of Maryland (UMD), specifically within the Department of Animal and Avian Sciences.

Conus Medullaris: Spinal Cord Anatomy Guide

Name: Conus Medullaris: Spinal Cord Anatomy Guide
Creator: Dr. Zubair Khalid

By Dr. Zubair Khalid, DVM, MS, PhD · 2026-09-20

The conus medullaris is the cone-shaped, tapered termination of the spinal cord, where the solid cord narrows and ends before continuing as the filum terminale. In dogs it typically ends near the L6 to L7 vertebral level, and the exact level varies with species, breed, body size and the individual animal.

That single anatomical fact decides where a clinician can safely place a needle. Every lumbar puncture, every lumbosacral epidural and every cerebrospinal fluid tap depends on knowing where the cord stops and where only nerve roots and meninges remain. A student who can picture the conus, the cauda equina and the filum terminale as three distinct structures will read spinal imaging, plan anesthesia and interpret neurologic deficits far more accurately than one who memorizes a single number.

This article is educational and is not a substitute for veterinary diagnosis or treatment.

What the Conus Medullaris Is

The spinal cord is the caudal continuation of the brainstem. It runs from the foramen magnum through the vertebral canal, giving off paired spinal nerves at each segment. Early in development the cord and the vertebral column are nearly the same length, and each spinal segment sits beside its matching vertebra. Growth then diverges. The vertebral column elongates faster than the cord, so the cord's caudal end shifts cranially within the canal and the lowest nerve roots must travel progressively farther to exit [1].

That differential growth produces the conus medullaris. The name is Latin for "medullary cone," and the structure is exactly that: the cord tapers over its final segments into a slender cone rather than ending in a blunt stump. The cone contains the sacral and coccygeal spinal segments, so it houses the lower motor neurons and reflex arcs for the bladder, the anal sphincter, the tail and the perineum.

Three structures sit at the caudal end of the cord and are constantly confused with one another.

Conus medullaris. The tapered terminal segment of the actual spinal cord tissue. It contains gray matter and functioning neurons.

Filum terminale. A thin, fibrous strand of pia mater that continues caudally from the tip of the conus and anchors the cord to the coccyx [2]. It carries no neurons. It is a tether, not a nerve.

Cauda equina. The bundle of nerve roots that surround the filum terminale below the conus. These are the dorsal and ventral roots of the lumbar, sacral and coccygeal spinal nerves, which must run caudally inside the dural sac before exiting at their intervertebral foramina. The name means "horse's tail," and the resemblance is genuine.

The relationship is simple once stated: the conus is the end of the cord, the filum is the string that hangs from it, and the cauda equina is the collection of roots that travels alongside both.

Summary Table: Structures at the Caudal Spinal Cord

Structure	Tissue type	Contains neurons?	Function	Clinical significance
Conus medullaris	Spinal cord (gray and white matter)	Yes	Sacral and coccygeal segments, bladder and sphincter reflexes	Injury causes conus medullaris syndrome with early, symmetric sphincter loss
Filum terminale	Pia mater, fibrous	No	Anchors cord to coccyx	Abnormal thickening or tethering produces tethered cord syndrome
Cauda equina	Nerve roots (dorsal and ventral)	Axons only, no cord gray matter	Conducts signals between cord and periphery	Injury causes cauda equina syndrome with asymmetric, root-pattern deficits
Dural sac	Meninges	No	Encloses cord, roots and cerebrospinal fluid	Its termination level determines where cerebrospinal fluid can be sampled

Vertebral Level of Termination Across Species

The conus medullaris does not end at a fixed vertebra. It ends at a level that reflects the relative growth of cord and column in that species. Comparative anatomy textbooks and imaging studies agree on the general pattern: the larger and longer the animal, the more cranial the conus sits relative to the total vertebral count, because the column outgrows the cord more dramatically.

Dogs. A large magnetic resonance imaging study of 117 dogs without spinal cord compression found that most breeds had their conus medullaris terminating at L6 [3]. The study also showed that heavier dogs had a more cranial termination, and that a small number of breeds deviated substantially. Cavalier King Charles Spaniels and Corgis reached the sacrum, while Boston Terriers ended more cranially at L5 or L6 [3]. Tail morphology and breed both influenced the level [3]. The practical summary for dogs is L6 to L7 as the usual range, with real breed variation in both directions.

Cats. The feline cord is shorter relative to the column than the dog's. Standard anatomy teaching places the feline conus in the region of L6 to L7 as well, with the cauda equina filling the sacral canal. Cats are small enough that the absolute distances are short, which is why feline lumbosacral epidurals demand precise technique.

Horses. The equine cord terminates further cranially in absolute terms but the horse has a long lumbar and sacral series, so the conus typically ends around the L5 to S1 region. The cauda equina in the horse is long and occupies much of the sacral canal, which is why sacral epidural anesthesia is feasible and why the cord itself is not at risk from a correctly placed caudal epidural.

Cattle. The bovine conus ends in the region of L6 to S1, with the cauda equina extending caudally into the sacral canal. As in the horse, the lumbosacral space is well below the cord, so lumbosacral epidural injection is a standard procedure in bovine practice.

Sheep. A computed tomography myelography study of 40 adult Merino-mixed sheep found that the conus terminated at S1 in 50 percent of animals and at S2 in the other 50 percent, with no terminations cranial to the lumbosacral space [4]. The dural sac extended caudal to the sacrococcygeal junction in 93.9 percent of sheep [4]. This is a useful reminder that even within one species the conus can sit a full segment lower than expected.

Humans, for comparison only. In people the conus typically ends at the lower third of the L1 vertebral body. A study of 1,000 adults found the most frequent location was the lower third of L1, with good interobserver agreement [5]. A separate study of 331 adults found a median termination at the lower third of L1 and reported that 97.9 percent of subjects had a conus at or above the L2 vertebra [6]. Another series of 944 adults found 18.9 percent had a conus at or below the L1-L2 interspace, and that spinal anesthesia at the L2-L3 interspace carried a 0.7 percent incidence of neuraxial risk [7]. These figures explain why human anesthesiologists target L3-L4 or lower. They are not transferable to quadrupeds, whose cord ends far more caudally relative to the lumbosacral space.

Species Comparison Table: Cord Termination and Cauda Equina Extent

Species	Typical conus medullaris level	Cauda equina extent	Practical note
Dog	L6 to L7, breed dependent	Through sacral canal to Cd1-2	Heavier dogs end more cranially [3]
Cat	L6 to L7	Through sacral canal	Short working distances, precise technique needed
Horse	L5 to S1	Long, fills sacral canal	Sacral epidural is safely below the cord
Cattle	L6 to S1	Extends into sacral canal	Lumbosacral epidural is standard practice
Sheep	S1 or S2	Caudal to sacrococcygeal junction	Dural sac extends beyond sacrococcygeal junction in most animals [4]
Human (comparison)	Lower third of L1	Through sacral canal	Target L3-L4 or lower for spinal anesthesia [6][7]

Why the Cord Ends Above the Column

The mechanism is differential longitudinal growth. During development the neural tube and the vertebral column are initially matched in length. The column then grows faster, and because the cord's cranial attachment at the foramen magnum is fixed, the extra column length is added caudal to the cord. The result is that the cord's terminal segments sit progressively more cranial to their corresponding vertebrae [1].

A second consequence follows from the same process. Each spinal nerve must still exit at its own intervertebral foramen, which is now some distance caudal to the cord segment that produced it. The lower lumbar, sacral and coccygeal roots therefore run obliquely and caudally inside the dural sac, forming the cauda equina. The more caudal the root, the longer its intradural course.

This is why the anatomy of the spinal cord cannot be taught as a simple stack of segments aligned with vertebrae. The cervical enlargement spans roughly C6 to T1 and the lumbar enlargement roughly L4 to S1 in many mammals, with the segment-to-vertebra offset increasing caudally [8]. A clinician who assumes that the L4 cord segment lies beside the L4 vertebra will misjudge lesion localization.

The Filum Terminale and Its Clinical Weight

The filum terminale is a continuation of the pia mater that extends from the conus tip to the coccyx [2]. It has two parts. The filum terminale internum lies inside the dural sac. The filum terminale externum continues beyond the dural sac to attach to the coccyx. A cadaveric study of 30 adult specimens found the conus most frequently terminated at the lower third of L1, the filum fused with the dural sac predominantly at the upper third of S2, and the mean total filum length was 221.83 millimeters with all segmental diameters below the 2 millimeter pathological threshold [2].

The filum is clinically important because it tethers the cord. If the filum is abnormally thick or short, or if it attaches abnormally low, the cord is held caudally and stretched as the vertebral column grows. This is tethered spinal cord syndrome, defined radiologically as a low-lying conus terminating at or below the lower third of the L2 vertebral body [9]. In a study of 105 children with arthrogryposis multiplex congenita, 40 percent had tethered cord, a significantly higher proportion than in the unaffected population, and 16 underwent detethering by filum sectioning [9].

Tethered cord matters in veterinary patients too. A low conus changes the safe window for neuraxial procedures. A case report of a 65-year-old human patient with tethered cord syndrome and a cord extending to L5-S1 described how the traditional L4-L5 puncture site would have placed the needle directly into cord tissue, and the anesthesia team chose general anesthesia instead [10]. The lesson transfers directly: an animal with a low conus has a smaller safe epidural window than its breed or size would predict.

Why Needle Sites Are Chosen Below the Conus

The logic of neuraxial anesthesia is spatial. A needle placed into the vertebral canal will contact whatever tissue lies in its path. Above the conus, that tissue includes the spinal cord. Below the conus, the canal contains only cerebrospinal fluid, nerve roots, meninges and the filum, none of which is cord parenchyma.

The human literature quantifies the risk. A study of 944 adults found that 18.9 percent had a conus at or below the L1-L2 interspace, and that spinal anesthesia at the L2-L3 interspace carried a 0.7 percent incidence of neuraxial risk [7]. A separate study of 331 adults found 97.9 percent had a conus at or above L2, and noted that with a needle misplaced one or two interspaces above the target, 2.1 percent of cords could be exposed at the L2-L3 interspace and 33.5 percent at the L1-L2 interspace [6]. Those numbers explain the classical human rule: puncture at L3-L4 or lower.

Veterinary practice follows the same principle with different landmarks. In dogs and cats, the lumbosacral space between L7 and S1 is the standard site for epidural injection and for cerebrospinal fluid collection, because the conus has already ended and the cauda equina is the only neural tissue present. In horses and cattle, the lumbosacral or sacrococcygeal space serves the same purpose. In sheep, the conus ends at S1 or S2 and the dural sac often extends beyond the sacrococcygeal junction, so lumbosacral or sacrococcygeal epidural injection still requires careful site selection to avoid iatrogenic cord injury [4].

The rule is not "puncture at a fixed vertebra." The rule is "puncture caudal to the conus in that individual." Imaging, breed knowledge and clinical judgment all feed into that decision.

How the Conus Is Identified in Practice

Magnetic resonance imaging. MRI is the reference standard for locating the conus in living patients. Observers divide each vertebral body into thirds and record which third contains the conus tip, a method with good interobserver agreement [5]. The same approach is used in the veterinary breed study that established L6 as the typical canine level [3].

Computed tomography myelography. Injecting iodinated contrast into the lumbosacral subarachnoid space opacifies the cerebrospinal fluid column and outlines the cord, the conus and the dural sac. This technique was used to map the ovine conus and dural sac termination [4].

Cadaveric dissection. Anatomical studies remove the dorsal laminae, open the meninges and record the vertebral level of the conus tip directly [1][2]. Fetal and cadaveric series provide normative data that imaging studies then test in living animals.

Neurologic examination. The conus contains the sacral segments. A lesion there produces early, symmetric loss of bladder and anal sphincter function, often with perineal sensory loss and a flaccid tail. A lesion of the cauda equina produces asymmetric, root-pattern deficits that may spare the sphincters until late. Distinguishing the two syndromes at the bedside is a core clinical skill [11].

Conus Medullaris Syndrome Versus Cauda Equina Syndrome

These two syndromes are the classic clinical pairing, and the distinction rests on anatomy. The conus is cord tissue, so a lesion there damages upper and lower motor neurons and the sacral reflex arcs together. The cauda equina is a bundle of roots, so a lesion there damages individual peripheral nerves.

A study of 87 patients with acute thoracolumbar spinal injuries classified lesions by the distance from the terminal end of the cord to the narrowest point of the spinal canal. The epiconus was defined as the region from the cord's terminal end to the proximal 1.0 to 2.25 vertebral bodies, the conus medullaris as the region proximal to less than 1.0 vertebral body, and the cauda equina as the distal nerve roots [12]. The narrowest levels were epiconus in 22 patients, conus in 37 and cauda equina in 25. On admission, significantly more patients with a narrowed epiconus had severe Frankel grades A to C than those with a narrowed cauda equina, but at final follow-up there was no significant difference in neurologic recovery among the three groups [12].

The practical translation for veterinary students is that lesion level within the caudal cord predicts the pattern of deficit, not necessarily the final outcome. Symmetric sphincter loss points to the conus. Asymmetric sciatic or pudendal deficits point to the cauda equina. Mixed pictures occur when both are involved.

Common Misconceptions

"The conus ends at the same vertebra in every dog." False. Most breeds end at L6, but heavier dogs end more cranially and some breeds reach the sacrum [3]. Breed and body weight both matter.

"The filum terminale is a nerve." False. It is a pia mater strand with no neurons [2]. It conducts nothing. It anchors.

"The cauda equina is part of the spinal cord." False. It is a bundle of nerve roots that have already left the cord. This is why cauda equina lesions produce peripheral nerve signs.

"Human vertebral levels apply to animals." False. The human conus ends near L1 [6][5], while the dog's ends near L6 to L7 [3]. Quadrupeds have a proportionally longer lumbar and sacral cord relative to the lumbosacral space.

"If the epidural is below the conus, the cord cannot be injured." Mostly true but not absolutely. A tethered cord can extend to L5-S1 and place cord tissue at a conventional puncture site [10]. Abnormal anatomy defeats normal rules.

"Age changes the conus level." In humans, age did not significantly affect conus termination in one large series [6], and a pediatric study found no significant age effect on conus or thecal sac position [13]. The level is largely set by growth, not by aging.

Quick Review

The conus medullaris is the cone-shaped terminal segment of the spinal cord, containing sacral and coccygeal neurons.
It ends near L6 to L7 in dogs, with breed and body weight shifting the level [3].
The filum terminale is a fibrous pia strand, not a nerve, and it anchors the cord to the coccyx [2].
The cauda equina is the bundle of lumbar, sacral and coccygeal nerve roots below the conus.
Needles go below the conus because only roots, meninges and cerebrospinal fluid remain there.
Conus lesions cause symmetric sphincter loss. Cauda equina lesions cause asymmetric root deficits [11].
A low-lying conus (tethered cord) shrinks the safe window for neuraxial procedures [9][10].

Clinical Relevance, Limitations and Common Mistakes

The conus medullaris is the single most important landmark for safe neuraxial access in every domestic species. Get the level wrong and the needle enters cord parenchyma. Get it right and the procedure is routine.

The most common mistake in practice is assuming a fixed vertebral level across all patients. The breed data show that heavier dogs end more cranially and that Cavalier King Charles Spaniels and Corgis can reach the sacrum [3]. A clinician who uses one number for every dog will eventually place a needle too close to cord tissue.

The second mistake is confusing conus and cauda equina syndromes on examination. Symmetric sphincter loss with a flaccid tail suggests the conus. Asymmetric sciatic or pudendal deficits suggest the cauda equina. The distinction changes the imaging plan and the prognosis discussion.

The third mistake is ignoring the dural sac. The dural sac can extend well beyond the conus, as in sheep where it reached caudal to the sacrococcygeal junction in 93.9 percent of animals [4]. Cerebrospinal fluid can therefore be sampled at a level where the cord has already ended, which is exactly why lumbosacral collection works.

A limitation worth stating plainly: individual anatomy varies, and no table substitutes for imaging or clinical assessment in a specific patient. Every case needs a veterinarian who can evaluate that animal's actual anatomy and clinical status.

Frequently Asked Questions

What is the conus medullaris?

The conus medullaris is the tapered, cone-shaped terminal segment of the spinal cord, containing the sacral and coccygeal spinal segments. It ends where the solid cord stops and the filum terminale begins.

Where does the spinal cord end in dogs?

In most dogs the conus medullaris ends near the L6 to L7 vertebral level. Heavier dogs tend to end more cranially, and some breeds such as Cavalier King Charles Spaniels and Corgis can reach the sacrum [3].

What is the difference between the conus medullaris and the cauda equina?

The conus medullaris is actual spinal cord tissue containing neurons. The cauda equina is the bundle of nerve roots that runs caudally from the cord below the conus. One is central nervous system, the other is peripheral nerve.

Why is epidural anesthesia placed below the conus medullaris?

Below the conus, the vertebral canal contains only nerve roots, meninges, cerebrospinal fluid and the filum terminale. There is no cord parenchyma to injure, so needle placement is safer.

What is the filum terminale?

The filum terminale is a thin fibrous strand of pia mater that extends from the tip of the conus medullaris to the coccyx. It contains no neurons and functions as an anchor for the cord [2].

What is tethered cord syndrome?

Tethered cord syndrome occurs when the conus terminates abnormally low, at or below the lower third of the L2 vertebral body, because the filum or another structure holds the cord caudally [9]. It can complicate spinal anesthesia because the cord occupies a lower position than expected [10].