Pathophysiology

Spinal nerve root compression commonly occurs in conditions such as acute herniated disc, spinal stenosis, trauma (e.g., burst fractures), metastatic or primary tumors of the spine, or spinal infections (e.g., epidural abscess) (Box 105-2). Acute CES most commonly presents secondary to lumbosacral intervertebral disc prolapse (Fig. 105-1). However, the pathophysiology of the symptoms and signs related to spinal nerve root compression remains poorly defined.

FIGURE 105-1 Thirty-year-old woman with acute cauda equina syndrome secondary to a large herniated disc at the L5-S1 level. The patient’s initial symptoms included urinary retention (postvoid residual = 200 mL), saddle anesthesia, and S1 motor-sensory radiculopathy. A, T2-weighted midsagittal MRI. B, Axial MRI demonstrating complete occlusion of the spinal canal at L5-S1. C, Axial MRI caudal to area of maximal compression (B) demonstrating displacement of the S1 roots and compression of the central sacral roots. Patient underwent urgent surgical (<24 hours of bladder symptoms) decompression with complete recovery of bladder function, sensation, and radiculopathy.

Several experimental studies have assessed the pathophysiologic mechanism of CES. Delamarter et al.^1,2 developed an animal model of CES, subjecting 30 beagle dogs to L6-7 laminectomy and cauda equina compression. Neurologic recovery was assessed in animals undergoing 75% constriction of the cauda equina followed by immediate, early, or delayed decompression. The first group was constricted and immediately decompressed. The remaining groups were constricted for 1 hour, 6 hours, 24 hours, and 1 week, respectively, before being decompressed. Evoked potentials were measured before and after surgery, before and after decompression, and 6 weeks after decompression. Six weeks after decompression, all dogs were killed, and the neural elements were analyzed histologically. After compression, all 30 dogs had significant lower extremity weakness, tail paralysis, and urinary incontinence. All dogs recovered significant motor function by 6 weeks after decompression. The dogs with immediate decompression typically recovered neurologic function within 2 to 5 days. The dogs receiving 1- and 6-hour compression recovered within 5 to 7 days. Dogs receiving 24 hours of compression remained paraparetic for 5 to 7 days, with bladder dysfunction persisting for 7 to 10 days and tail dysfunction for up to 4 weeks. The dogs with compression for 1 week were paraparetic and incontinent during the duration of cauda equina compression. They recovered the ability to walk by 1 week and regained bladder and tail control by the time of euthanasia. Immediately after compression, all five groups demonstrated at least 50% deterioration of the posterior tibial evoked potential amplitudes.² Delamarter et al.² demonstrated axoplasmic flow blockade and wallerian degeneration of the motor nerve roots distal to the constriction and of the sensory roots proximal to the site of constriction, as well as dorsal column degeneration. Severe arterial narrowing occurred at the level of the constriction with venous congestion of the roots and dorsal root ganglia of the seventh lumbar and first sacral nerves.¹ Evoked potentials were the most sensitive predictor of neural compression, revealing neurologic abnormalities before the appearance of neurologic signs and symptoms.¹ Cystometrograms were not sensitive until severe compression was achieved. Bladder dysfunction was correlated with axoplasmic flow blockade and early sensory changes during neurovenous congestion.

Olmarker et al.^3–7 developed an experimental model of acute, graded compression of the cauda equina in pigs that accurately mimics the neural and vascular anatomy of the human cauda equina. There were structural and vascular differences between spinal nerve roots and peripheral nerves that could contribute to differences in compression susceptibility between these two parts of the nervous system. Pressure transmission from the balloon to the nerve roots permitted determination of occlusion pressures for the arterioles, capillaries, and venules of the cauda equina.⁷ Arteriolar blood flow ceased when the applied pressure approached the mean arterial blood pressure. Capillary blood flow was dependent on flow in connected venules, and the blood flow in some venules ceased at 5 to 10 mm Hg despite venous occlusion pressures ranging from 5 to 60 mm Hg. Compression up to 200 mm Hg for 2 hours did not induce a no-reflow phenomenon upon compression release. However, transient hyperemia was noted at all pressure-time relations studied, indicating nutritional deficit in the compressed segment during compression. Signs of edema were observed in nerve roots exposed to compression for 2 hours at either 50 or 200 mm Hg. The nutritional supply to the cauda equina was impaired at low pressure levels (<10 mm Hg).^4,5 Thus diffusion from adjacent tissues with a better nutritional supply, including the cerebrospinal fluid, could not compensate completely for compression-induced effects on the transport of nutrients. However, a certain nutritional supply to the compressed segment was present even at 200 mm Hg compression. A rapid compression rate resulted in more pronounced effects on the nutritional supply than did a slow compression rate. Nutritional impairment was observed both within and outside the compressed nerve segment. An increase in vascular permeability was induced by compression at 50 mm Hg for 2 minutes.⁶ The magnitude of this permeability increase was dependent on both the magnitude and the duration of compression. The permeability increase was more pronounced for the rapid compression onset rate than for the slow compression onset rate at all pressure-time relations studied. Reduction of muscle action-potential amplitude in tail muscles, after stimulation cranial to the compression zone, was induced by compression at 100 and 200 mm Hg for 2 hours.^4,6

Pedowitz et al.⁸ and Rydevik et al.^9,10 presented an experimental model of compression-induced functional changes of the porcine cauda equina that permits electrophysiologic investigation of the neurophysiologic changes induced by nerve root deformation. In several studies, they compared the effects of various pressures and durations of acute compression on spinal nerve root conduction in the pig cauda equina. Changes in both afferent (compound nerve action potentials) and efferent (compound motor action potentials) conduction were induced at an acute pressure threshold of 50 to 75 mm Hg. Higher compression pressures produced a differential recovery in afferent and efferent conduction.^9,10 Efferent conduction and afferent conduction were monitored during compression for 2 or 4 hours with compression pressures of 0 (sham treatment), 50, 100, or 200 mm Hg. Recovery was monitored for 1.5 hours. No significant deficits in spinal nerve root conduction were observed with 0 or 50 mm Hg compression, whereas significant deficits were induced by 100 and 200 mm Hg compression. Variance analysis demonstrated significant effects of compression pressure and duration on conduction, with a significant difference between efferent and afferent conduction at the end of the recovery period, suggesting a synergistic interaction between biomechanical and microvascular mechanisms in the production of nerve root conduction deficits.⁸

Compression of the spinal nerve roots often occurs at multiple levels simultaneously; however, the basic pathophysiology of multilevel compression is poorly defined. Using a thermal diffusion technique, Takahashi et al.¹¹ quantitated intraneural blood flow in the uncompressed segment between two compressive balloons in the porcine cauda equina. At 10 mm Hg compression, there was a 64% reduction of total blood flow in the uncompressed segment compared with precompression values. Total ischemia occurred at pressures 10 to 20 mm Hg less than the mean arterial blood pressure. After two-level compression at 200 mm Hg for 10 minutes, there was a gradual recovery of the intraneural blood flow toward the baseline. Recovery was less rapid and less complete after 2 hours of compression. Double-level compression of the cauda equina induced blood flow impairment, not only at the sites of compression but also in the intermediate nerve segments located between two compression sites, even at very low pressures.

Cauda Equina Syndrome Secondary to Disc Herniation

Diagnostic Imaging

Diagnosis and treatment are often delayed because of lack of recognition of the condition and failure to appreciate the surgical imperative for its treatment. Once cauda equina compression is recognized or suspected, MRI is the investigation method of choice. If an MRI is contraindicated, of poor quality because of motion artifact (patients with CES are often in severe pain), or unavailable, CT-myelography is recommended. CT alone can be misleading in cases of complete canal occlusion (Fig. 105-2).

FIGURE 105-2 A, Axial CT at L4-5 of a 43-year-old male patient with right L5 radiculopathy and abnormal perineal sensation. Axial (B) and sagittal (C) MRIs of the same patient. As can be seen, the contrasting effect of epidural fat or cerebrospinal fluid in CT can be lost because of massive disc herniations causing near or complete occlusion of the spinal canal. This patient’s CT was interpreted as equivocal for an L4-5 disc herniation. Consequently, MRI and CT-myelography are the imaging modalities of choice for the evaluation of cauda equina syndrome.

Surgical Therapy

Urgent surgical intervention should commence after diagnosis. Choudhury and Taylor²⁰ advocated wide laminectomy with excision of the overhanging facet joints and adequate visualization of the lumbar nerve roots. This yielded good or excellent results in 95% of patients and fair results in the remainder. No postoperative spinal instability or significant morbidity was reported.²⁰

A routine microdiscectomy interlaminar approach necessitates retraction on already severely compromised nerve roots. Furthermore, because of the extent of compression, these roots often have no available canal space; thus, retraction not only increases traction but also may cause direct compression (against the disc or the overlying lamina). For the same reason, the use of large punches with a thick foot plate is also not recommended (compresses underlying roots against the herniated disc) (Fig. 105-3). Consequently, the authors advocate a wide bilateral decompression (laminectomy and medial third facetectomy). To avoid iatrogenic compression of the cauda equina, the surgeon can perform the laminectomy using a high-speed bur to thin out the lamina and then curettes and small punches lateral to the area of maximal compression to release the medial portion of the lamina (see Fig. 105-3). Once the lamina is released, it can be safely lifted away from the cauda equina. The decompression should be adequate enough to allow access lateral to the thecal sac and traversing roots at the level of the affected disc space and/or the sequestered fragment. The surgeon should carefully perform retrieval of the herniated fragment to avoid creating a further mass effect and therefore increased traction on the less mobile central sacral roots. This can be accomplished by performing a lateral anulotomy and discectomy and then pushing the central fragment back into the disc space with a reverse-angle curette. If the fragment is sequestered, it should be manipulated in a lateral direction using a nerve hook or angled curette and then retrieved. Before closure, the surgeon should confirm adequate decompression.

FIGURE 105-3 Sagittal (A) and L5-S1 axial (B) MRIs of a 45-year-old woman with cauda equina syndrome demonstrating severe compression of the cauda equina. C, The cauda equina (outlined in white) is draped over the disc and compressed between the disc and lamina. Removal of the lamina (medial to the black arrows) by sublaminar placement of standard spinal punches (see text) would obviously cause further compression of the roots. Wide bilateral laminectomy is required to gain safe access to the disc. Following appropriate decompression of the cauda equina, care must be taken when retrieving large disc fragments to avoid excessive traction or mass effect on the tented traversing nerve roots.