Conus Medullaris Syndrome

The conus medullaris is the tip of the spinal cord. In adults it lies at the L1-L2 disc in 65 to 75 percent of people, with a normal range from T12 to L3, and a lesion there produces a syndrome of early, symmetric sphincter dysfunction with saddle anaesthesia and a mixture of upper and lower motor neurone signs. It is a surgical emergency: outcome depends on the cause, on how complete the sphincter loss is at presentation, and on the time to decompression.

Why it is missed. The patient often has minimal back pain, so the diagnosis is delayed, and a delayed conus lesion is a recurring source of complaints and litigation worldwide. The cognitive trap is to label every acute sphincter disturbance a "cauda equina syndrome"; the cord-tip lesion behaves differently and warrants equally urgent imaging. Bladder retention or incontinence, loss of anal tone and erectile dysfunction are often the presenting features, sometimes before overt motor weakness is recognised.

Causes. The commonest structural causes each have their own page.

• Trauma: burst fractures at the thoracolumbar junction are the classic cause
• Tumour: metastases (breast, lung, prostate, kidney, lymphoma) and primary tumours (ependymoma, astrocytoma, haemangioblastoma)
• Disc herniation: a large central disc at L1-L2 can compress the conus
• Tethered cord, lipoma, dermoid, cyst: relevant in younger patients (tethered cord syndrome)
• Ischaemia: anterior spinal artery infarct at the conus

What it costs the patient. Bladder and bowel dysfunction dominate the picture and persist even after surgery, with sexual dysfunction in men alongside them (the figures are in the complications table). Permanent saddle anaesthesia impairs continence, hygiene and seating comfort, and loss of sensation combined with immobility invites pressure sores. The psychosocial burden of depression, loss of independence and relationship strain follows.

The anatomy. The spinal cord ends as the conus medullaris. Within it sit the sacral cord segments S2 to S5 and the coccygeal segments, which supply the perineum, bladder, bowel and sexual function. The lumbosacral enlargement above (L1-S2) gives rise to the lumbar and sacral nerve roots, and those roots (L2-S5) descend below the conus as the cauda equina to exit at their respective neural foramina.

View larger

Why the signs are mixed. A lesion at the conus damages both the cord tracts passing through it and the sacral anterior horn cells or roots at its tip. The corticospinal tracts give the upper motor neurone signs of spasticity, hyperreflexia and Babinski; the anterior horn cells give the lower motor neurone signs of flaccid weakness, wasting and areflexia. Both patterns coexist, which is the signature of a cord-tip lesion.

Why sphincter loss is early. The sacral segments S2 to S4 carry the parasympathetic fibres to the detrusor, and the pudendal nerve (S2-S4) controls the external urethral and anal sphincters. A compressive lesion at the conus interrupts both the preganglionic and the somatic outflow at once, so the detrusor becomes areflexic, with retention and then overflow incontinence, and anal tone is lost.

Why the pattern is symmetric. Midline compression of the conus affects both halves of the cord equally, the saddle area is innervated by the S2 to S5 dermatomes of both sides, and both pudendal nerves are compromised together, so the sensory and sphincter loss is bilateral. In the cauda equina the roots are lateralised within the thecal sac, which is why cauda lesions are often patchy.

Conus lesions are classified on two axes: what caused them and how complete they are.

History. The tempo of onset points to the cause: acute with trauma or disc, subacute with tumour, chronic with an intramedullary lesion. Ask about each domain in turn.

• Pain: back pain, bilateral leg pain, perineal discomfort
• Sensory: saddle numbness of the perineum, buttocks and posterior thighs, often bilateral
• Bladder: retention with overflow, urgency, frequency, incontinence
• Bowel: constipation, loss of awareness of rectal fullness
• Sexual: erectile dysfunction, loss of ejaculation, loss of genital sensation

Examination. The classic finding is brisk knee jerks with absent ankle jerks: the knee jerk (L3-L4) is often preserved and the ankle jerk (S1) often absent. A Babinski sign or ankle clonus marks the corticospinal tracts as involved.

• Motor: test the L2-S1 myotomes and record whether weakness is proximal or distal
• Reflexes: knee jerk, ankle jerk, plantar responses, clonus
• Sacral reflexes: anal wink (S2-S4) and bulbocavernosus (S2-S4), always
• Per rectal examination: anal tone (a patulous anus), voluntary squeeze, and a post-void residual
• Sensory: map the deficit over the perineum, buttocks and posterior thighs

Conus or cauda equina. The conus is a cord lesion at L1-L2 and the cauda equina is a root lesion below it, and the examination separates them on symmetry, the timing of sphincter loss, the prominence of pain and the presence of upper motor neurone signs. The distinction does not change the urgency of imaging; it changes what the imaging must include.

The other differentials. Once conus and cauda equina have been separated, acute sphincter dysfunction has a short list of other causes.

Sacral sparing and completeness. Whether a cord-tip injury is complete or incomplete is defined by sacral sparing, assessed at exactly the most caudal segments the conus contains. The ISNCSCI examination tests three components, and the presence of any of them makes the injury incomplete:

1. Light touch and pin-prick sensation at S4-5
2. Deep anal pressure (DAP) felt on rectal examination
3. Voluntary anal contraction (VAC)

An injury is ASIA Impairment Scale A (complete) only when there is no motor or sensory function in the lowest sacral segments; any sacral sparing makes it at least AIS B (sensory incomplete) and carries a substantially better prognosis for recovery. The wider grading framework is set out in spinal cord injury.

The spinal-shock caveat. An injury cannot reliably be called complete until spinal shock has resolved, and the return of the bulbocavernosus reflex signals that resolution; only after it does a persistent absence of sacral sparing genuinely indicate a complete injury. That caveat interacts with the surgical-timing debate below: a patient labelled "complete" in the first hours may not be, which is precisely why an elapsed clock is a poor reason to decline decompression, and why Gleave's nihilism about the established complete syndrome cannot safely be applied to the acute presentation.

MRI of the whole spine with contrast is the first-line investigation, and the emphasis on the whole spine is deliberate. Skip lesions occur, and the lesion responsible for conus-type signs may sit above the lumbar region: a patient with saddle anaesthesia and a negative lumbar MRI may harbour a dorsal cervicothoracic lesion such as a dural AV fistula, or a spinal epidural abscess at another level. Even when a lumbar disc is suspected, the imaging must include the conus tip at L1-L2.

• Sequences: sagittal T2, axial T2 through the conus, T1 before and after gadolinium
• Look for: cord expansion, conus oedema, intramedullary lesion, epidural mass, disc, haematoma, abscess, fracture fragment
• Gadolinium improves detection of tumour, abscess and leptomeningeal disease

In trauma, CT plus MRI of the thoracolumbar junction is mandatory when neurological signs are present, even with a normal lumbar MRI.

CT myelography is the alternative when MRI is contraindicated by a pacemaker, severe claustrophobia or metal. Intrathecal contrast is injected at lumbar puncture and the spine is then scanned, looking for a block at the conus level, mass effect and, indirectly, cord expansion. It is invasive and carries the risk of post-dural-puncture headache.

The bladder at the bedside. A bladder scan after voiding with a post-void residual of more than 200 mL suggests retention. Insert a urethral catheter early, both to protect the upper tracts and to document the retention. Urodynamics assess detrusor areflexia, sphincter dyssynergia and capacity, and they carry prognostic weight: complete areflexia predicts poorer bladder recovery.

Spinal angiography or CT angiography is reserved for a suspected vascular malformation, typically a dural AV fistula, and shows enlarged perimedullary veins and early venous filling. Endovascular embolisation is first line for many dural AV fistulas, and recovery is better if the fistula is treated before the cord damage becomes irreversible.

Typing the neurogenic bladder. The bladder is the dominant disability, and its behaviour depends on where the lesion sits relative to the sacral cord. A suprasacral lesion, cord injury above the conus, leaves a spastic, hyperreflexic ("automatic") bladder: the detrusor contracts reflexively, often with detrusor-sphincter dyssynergia (DSD), generating high intravesical pressures that threaten the upper tracts with reflux and hydronephrosis. A conus or cauda lesion destroys the sacral parasympathetic outflow (S2-4) and pudendal control, so the detrusor does not contract: a flaccid, areflexic ("autonomous") bladder with retention, overflow incontinence and a large post-void residual.

Why the type dictates the treatment. The flaccid lower motor neurone bladder is emptied by clean intermittent self-catheterisation (CISC), whereas the spastic upper motor neurone bladder needs anticholinergics or antimuscarinics, or onabotulinumtoxinA, to lower its storage pressure. Whatever the type, the goal is low-pressure storage and complete emptying to protect the upper urinary tracts; renal failure was historically a leading cause of death in spinal cord injury, which is why urology follow-up and urodynamics are essential.

What decides the outcome. Cause, completeness at presentation and time to decompression, and the table sets out which way each cuts. Earlier surgery is associated with better sphincter recovery, but the conventional 24 to 48 hour benchmark is contested for the reasons set out in the management section.

Read the timing papers as a live argument, not a consensus. Two of the four below (Gleave, Chau) are sceptical of the 48-hour rule and two (Ahn, DeLong) supply the quantitative case for early surgery — and even those two find no advantage to operating inside 24 hours, and confidence intervals that often cross 1. Grouped by what each actually answers.