Spinal Cord Infarction

Published on 12/04/2015 by admin

Last modified 22/04/2025

Print this page

This article have been viewed 5057 times

Chapter 9 Spinal Cord Infarction

Alejandro A. Rabinstein, Steven J. Resnick

Spinal cord infarction remains the less well-studied form of acute ischemic stroke. Although several useful studies describing the clinical and radiological features of spinal cord infarction have been published and the characteristics of the presenting syndrome are fairly well known, its risk factors (apart from aortic dissection and aortic surgery), area of maximal cord involvement, and prognosis are insufficiently understood. Furthermore, it remains a condition with no proven effective treatment.

Magnetic resonance imaging (MRI) allows documentation of spinal cord infarction. However, the true sensitivity of MRI for the early diagnosis of spinal cord ischemia is not well established. Imaging the cord is also important to exclude other causes of acute spinal cord syndrome such as cord compression (from displaced discs, epidural hematomas or abscesses, intradural extramedullary tumors, etc.), spinal cord hemorrhage, dural arteriovenous fistula, multiple sclerosis, neuromyelitis optica, infectious myelitis (e.g., West Nile virus), transverse myelitis (most often negative on early imaging), spinal cord contusion, or intramedullary tumors.¹^–³ MRI cannot be replaced by any other imaging modality for assessment of spinal cord infarction and exclusion of its differential diagnoses.

VASCULAR ANATOMY OF THE SPINAL CORD

The spinal cord is supplied by three main arteries: the anterior spinal artery and the two smaller posterior spinal arteries. The anterior spinal artery is located along the ventral midline of the cord and supplies the anterior two thirds (often closer to the anterior 75%) of the cord tissue. The posterior spinal arteries lie on each side of the posterior aspect of the cord and supply its posterior third (or 25%). Thus the anterior, central, and lateral regions of the cord are irrigated by the anterior spinal artery and the dorsal horns and columns receive blood from the ipsilateral posterior spinal artery. An internal watershed area can be found in the central cord between small penetrating branches from the anterior and posterior spinal arteries. Branches from the three spinal arteries encircle the surface of the cord forming a fine pial plexus with multiple anastomosis (known as the vaso corona).

The anterior spinal artery receives blood from the vertebral arteries in the cervical region and from radicular arteries in the thoracic and lumbar regions. Often two branches arising from each vertebral artery join at the upper cervical level to form the anterior spinal artery; however, many anatomical variations exist and these branches may arise from the posterior inferior cerebellar arteries or from cervical segmental branches. Just a few radicular arteries are responsible for the blood supply to the spinal arteries at the thoracolumbar level. They stem from segmental branches of the aorta (posterior intercostal and lumbar branches), which reach the intervertebral foramina and divide into the anterior and posterior radicular arteries. The largest radicular artery is the arteria radicularis magna of Adamkiewicz (or main anterior radicular artery), which most commonly arises on the left from T9 to T12 but occasionally can be positioned on the right (17% of cases) and arise anywhere from T5 to L4. Between the lower cervical and the mid- to lower thoracic levels, there are usually just two or three small radicular branches supplying this long segment of the cord. Hence the midthoracic area is traditionally considered a watershed territory at high risk for ischemia from hypoperfusion.^4,⁵ Yet most cases of documented spinal cord ischemia do not occur in this area.⁶^–⁸

Figures 9-1 and 9-2 illustrate the normal vascular anatomy of the spinal cord.

Figure 9-1 Longitudinal view of the normal vascular anatomy of the spinal cord.

Figure 9-2 Axial view of the normal vascular anatomy of the spinal cord.

Case Vignette

A 52-year-old woman with history of ovarian cancer in remission and no previous vascular disease presented with sudden onset of bilateral leg weakness. She had acute back pain initially, but it was short lasting. There was no loss of sphincter control. In the emergency department, she had flaccid paraplegia and leg areflexia. On sensory examination, she had decreased superficial pain and temperature sensation in both legs, but there was no sensory level detected in the trunk. Proprioception was normal. Computed tomography (CT) angiogram of chest and abdomen excluded aortic dissection but disclosed diffuse atherosclerotic changes throughout the descending aorta. MRI of the thoracolumbar spine revealed changes consistent with acute spinal cord ischemia from T11 to the tip of the conus medullaris (Figure 9-3). She was treated with interventions to optimize blood pressure, intravenous dexamethasone, and lumbar drainage. Over the following 48 hours, her sensation in the legs became nearly normal, but she only regained minimal motor function (some activation of hip flexors). She developed signs of neurogenic bladder and bowel. Comprehensive vascular evaluation uncovered no other mechanisms for the spinal cord ischemia other than aortic atherosclerosis. Three months later, her neurological condition remained unchanged.

Figure 9-3 Magnetic resonance imaging (MRI) scan of the thoracolumbar spine showing signs of acute spinal cord ischemia. Sagittal T2 fast spin echo sequence (upper left panel) demonstrates hyperintense signal and cord enlargement in the affected area (T11 to conus). Axial cut at the T12 level shows high signal intensity in the territory of the anterior spinal artery, more prominent in the ventral gray matter (upper right panel). Diffusion-weighted imaging (DWI) confirms the acuity of the lesion by revealing diffusion restriction (lower left panel). Post-gadolinium T1 fat-saturated sequence shows modest enhancement of the lesion on a repeat MRI scan 5 days after the onset of deficits (lower right panel).

♦ The presentation of spinal cord infarction depends on the underlying cause, the level of ischemic injury, and the extent of the infarction.

♦ Multiple risk factors and etiologies have been associated with the occurrence of spinal cord ischemia.^2,⁴ Aortic disease (atherosclerosis, dissection) and aortic surgery (repair of thoracoabdominal aortic aneurysm) are the most common causes of spinal cord ischemia.⁴^–⁹ Severe hypotension, especially that due to cardiocirculatory arrest, may also produce spinal cord ischemia,^4,^7,⁸ but this complication is infrequent. Presence of multiple vascular risk factors (hypertension, smoking, hypercholesterolemia, diabetes mellitus, previous vascular events) appears to increase the risk of spinal cord infarction.^6,⁷ Many cases remain classified as cryptogenic.^4,⁶^–⁹

♦ Anterior spinal artery infarction, exemplified by the previous vignette, is the most frequent form of spinal cord ischemia.^4,^6,^7,⁹ It typically presents with sudden flaccid paralysis (paraplegia or quadriplegia according to the level of the lesion), areflexia, loss of spinothalamic sensory modalities (pain and temperature), and autonomic deficits (such as atonic urinary bladder, paralytic ileus, and abolished sphincter tone) below the level of the lesion.⁴ Posterior column sensory modalities (vibration and proprioception) are preserved, resulting in dissociated sensory loss. Back or radicular pain may be severe at the site of infarction, but this is an inconsistent symptom. After the phase of acute spinal shock, pyramidal signs (spasticity, hyperreflexia, Babinski signs, clonus) supervene.

♦ Other clinical syndromes are encountered more infrequently. Posterior spinal artery territory infarction causes selective loss of proprioception or of all sensory modalities (due to dorsal horn involvement) with relative preservation of motor function.¹⁰ Deficits are often asymmetric. Posterior spinal artery ischemia is uncommon because of the extensive collateral network in the dorsal cord. Transverse cord infarction presents with flaccid weakness, and loss of all sensory modalities and autonomic control below the level of the lesion. Local pain or hyperesthesia is often reported at the level of the infarction. When unilateral or asymmetric, it may resemble the Brown-Sequard syndrome. This pattern of infarction is seen after severe systemic hypotension or spine trauma. Similar causes may result in centrospinal infarction, which affects the cord gray matter and manifests with persistent lower motor neuron weakness and dissociated loss of pain and temperature in a segmental distribution. When the injury is partially reversible (often seen with trauma), the legs recover motor function and the arms remain weaker distally more than proximally in a pattern that is often referred to as “man-in-the-barrel” syndrome.

♦ Ischemia most often involves the thoracolumbar region, followed by the cervical area (often in its higher portion).⁶^–⁸ Contrary to what would be expected on the basis of purely anatomical grounds, the epicenter of the infarction is rarely in the midthoracic region.

♦ MRI signs of spinal cord ischemia are best seen on T2-weighted sequence. It shows “pencil-like” hyperintensities on sagittal views, often associated with cord enlargement (see Figure 9-3, upper left).⁸ Preferential involvement of the ventral gray matter often gives the appearance of “owl eyes” on axial views (see Figure 9-3, upper right).^2,¹¹ Changes may be asymmetric, but they are typically bilateral.⁸ Increased bone marrow signal may indicate concurrent bone infarction.^12,¹³

♦ T1-weighted images are rarely diagnostic. When visible, lesions have decreased signal intensity. Subtle hemorrhagic components have been infrequently reported.⁸ Enhancement on post-gadolinium images may appear early but are most commonly visible a few days after the onset of symptoms (see Figure 9-3, lower right).

♦ Diffusion-weighted imaging (DWI) may be a valuable addition to T2-weighted sequence in the detection of acute spinal cord ischemia (see Figure 9-3, lower left). In most cases with restricted diffusion, T2 hyperintensities are also seen,¹⁴ but a few cases with diffusion abnormality in the absence of T2 changes have been reported.^15,¹⁶ The shortest interval between onset of symptoms of spinal cord ischemia and documented diffusion abnormalities reported thus far has been 3 hours (later than in brain ischemia). Pseudo-normalization of the apparent diffusion coefficient (ADC) occurs after only 7 days (earlier than in brain ischemia).¹⁷ Multishot, interleaved echo planar imaging may be less susceptible to image distortion and offer better spatial resolution than single-shot echo planar imaging and fast spin echo DWI.^14,^18,¹⁹

♦ The sensitivity of MRI for the diagnosis of spinal cord ischemia is not well established. In a consecutive series of 54 patients with acute spinal cord syndrome who underwent MRI within 45 days of symptom onset (median, 1 day), imaging was diagnostic in only 45% of cases.⁷ However, DWI was obtained in only five of the patients, and many of the cases included in this series had mild deficits at presentation. In another study of 28 consecutive patients examined with MRI (not including DWI) within 10 days of clinical onset, MRI was diagnostic in 24 cases (86%).⁶ Delayed MRI remained negative even after contrast administration in the four cases with lack of ischemic changes on initial MRI. Other published series did not examine patients consecutively, and therefore selection bias could explain the high sensitivity of MRI in these reports.^8,^9,^14,¹⁶ In our experience, the sensitivity of MRI is often seriously compromised by motion, pulsatility, and susceptibility artifacts.

♦ Angiography is usually not performed in patients with suspected or documented spinal cord ischemia, unless it is presumed to be related to a vascular malformation (dural arteriovenous fistula). Selective catheterization of radicular arteries is technically difficult and may be complicated with vessel injury or occlusion. Noninvasive angiography (most often magnetic resonance angiography) is sometimes used to exclude concomitant vascular anomalies,²⁰ but the diagnostic accuracy of this study remains undetermined.

Case Vignette

A 74-year-old man with history of severe peripheral vascular disease requiring bilateral iliac-to-femoral artery bypass surgery was admitted to the hospital for surgical repair of an enlarging thoracoabdominal aneurysm. A cerebrospinal drainage catheter was inserted before the initiation of the vascular surgery. The surgery was completed without apparent complications. However, following surgery the patient was noticed to be paraplegic. Neurological examination showed dense flaccid paraparesis with minimal movement of the toes as the only preserved motor function. No definitive sensory level was noted. MRI of the spine on postoperative day 2 was severely limited by motion artifact. A repeat MRI of the spine obtained 10 days after surgery (Figure 9-4) confirmed the presence of spinal cord infarction extending from T10–11 to the conus. The patient experienced no neurological recovery and was discharged 2 weeks later with persistent paraplegia, neurogenic bladder and bowel, and medications to control his severe back pain. He died 2 months after discharge in a nursing home from suspected pulmonary embolism (his preventive regimen of anticoagulation had been stopped after an episode of melena).

Figure 9-4 Sagittal T2-weighted fast spin echo sequence showing high-intensity signal extending from T10–11 to the conus; severe scoliosis only allows partial visualization of the length of the thoracic cord (left panel). Axial cut demonstrates predominant involvement of the anterior cord (right panel).

♦ Postoperative paraplegia from spinal cord ischemia is a well-recognized complication of thoracic aortic surgery, with an incidence between 10% and 30% depending on the type and emergency of the surgery.^5,²¹

♦ Preoperative spinal angiography, intraoperative evoked potentials, and perioperative use of lumbar drains may decrease the risk of spinal cord ischemia.^5,²¹^–²⁴

♦ Cerebrospinal fluid drainage is the best studied strategy for the prevention of paraplegia after surgical repair of thoracic and thoracoabdominal aneurysms. In a meta-analysis of randomized controlled studies and cohort studies with a control group, the pooled odds ratio for development of paraplegia in the cerebrospinal drainage group was 0.3 (i.e., 70% reduction in the risk of the neurological complication).²¹ Cerebrospinal fluid drainage can be combined with other preventive strategies such as regional spinal cord hypothermia with epidural cooling,²⁵ administration of naloxone,²² or distal aortic perfusion.²⁴ Extended duration of cerebrospinal fluid drainage (for up to 100 hours) may confer additional protection.²³ Hemorrhagic complications related to the placement of the lumbar drain may occur rarely, and they may be extremely difficult to recognize promptly.²⁶

♦ When patients are paraplegic upon awakening following aortic surgery, MRI can confirm the diagnosis in a timely manner and allow for adequate management of blood pressure to maximize perfusion to the cord.

MANAGEMENT AND PROGNOSIS

♦ Unfortunately, radiological confirmation of the diagnosis of spinal cord ischemia does not open major therapeutic opportunities. In postsurgical cases, as mentioned earlier, blood pressure should be supported to maximize cord perfusion. In nonperioperative cases, the options are even more limited. When the systemic blood pressure is low, it should be improved. Isolated cases of improvement of initially severe deficits after emergency placement of a lumbar drain have been reported,²⁷ and we tend to try this intervention when the diagnosis is made shortly after symptom onset. The value of corticosteroids has not been tested; although large doses of dexamethasone are sometimes administered extrapolating an approach validated for acute spinal cord trauma, we believe there is no solid rationale for the use of steroids in patients with nontraumatic spinal cord infarction.

♦ It is clear that more severe deficits at presentation portend worse long-term prognosis for functional recovery.^4,^6,⁷ In particular, patients with complete or nearly complete motor and sensory loss (American Spinal Injury Association [ASIA] Grades A and B), bladder dysfunction, or abolished proprioception in addition to paraplegia (indicating complete transverse ischemia) predict unfavorable long-term outcome.^6,⁷ However, the prognosis of spinal cord infarction in patients with more benign presentations may be much more favorable.^7,⁹ Even some patients with severe impairment at onset but without signs of complete transverse cord infarction may achieve substantial recovery.⁶ Other than initial severity of deficits, prognostic factors are unknown. In particular, the impact of imaging findings on prognosis for recovery remains to be adequately studied.

References

1 Brinar VV, Habek M, Brinar M, Malojcic B, Boban M. The differential diagnosis of acute transverse myelitis. Clin Neurol Neurosurg. 2006;108:278-283.

2 Fortuna A, Ferrante L, Acqui M, Trillo G. Spinal cord ischemia diagnosed by MRI. Case report and review of the literature. J Neuroradiol. 1995;22:115-122.

3 Fukui MB, Swarnkar AS, Williams RL. Acute spontaneous spinal epidural hematomas. AJNR Am J Neuroradiol. 1999;20:1365-1372.

4 Cheshire WP, Santos CC, Massey EW, Howard JFJr. Spinal cord infarction: etiology and outcome. Neurology. 1996;47:321-330.

5 Shamji MF, Maziak DE, Shamji FM, Ginsberg RJ, Pon R. Circulation of the spinal cord: an important consideration for thoracic surgeons. Ann Thorac Surg. 2003;76:315-321.

6 Masson C, Pruvo JP, Meder JF, Cordonnier C, Touze E, De La Sayette V, et al. Spinal cord infarction: clinical and magnetic resonance imaging findings and short term outcome. J Neurol Neurosurg Psychiatry. 2004;75:1431-1435.

7 Nedeltchev K, Loher TJ, Stepper F, Arnold M, Schroth G, Mattle HP, et al. Long-term outcome of acute spinal cord ischemia syndrome. Stroke. 2004;35:560-565.

8 Weidauer S, Nichtweiss M, Lanfermann H, Zanella FE. Spinal cord infarction: MR imaging and clinical features in 16 cases. Neuroradiology. 2002;44:851-857.

9 Novy J, Carruzzo A, Maeder P, Bogousslavsky J. Spinal cord ischemia: clinical and imaging patterns, pathogenesis, and outcomes in 27 patients. Arch Neurol. 2006;63:1113-1120.

10 Mascalchi M, Cosottini M, Ferrito G, Salvi F, Nencini P, Quilici N. Posterior spinal artery infarct. AJNR Am J Neuroradiol. 1998;19:361-363.

11 Mawad ME, Rivera V, Crawford S, Ramirez A, Breitbach W. Spinal cord ischemia after resection of thoracoabdominal aortic aneurysms: MR findings in 24 patients. AJNR Am J Neuroradiol. 1990;11:987-991.